Quality Improvement Tool Instruction Guide Diabetes Care

Transcription

1 PRIMIS: Quality Improvement Tool Instruction Guide Quality Improvement Tool Instruction Guide Diabetes Care PRIMIS development of the Diabetes Quality Improvement tool has been funded by Boehringer Ingelheim. Boehringer Ingelheim has undertaken a medico legal review but PRIMIS has retained editorial control and intellectual property rights for this tool. Prepared by PRIMIS August 2017 The University of Nottingham. All rights reserved. UK/DIB a(2) Diabetes Care Audit Diabetes_Care_V3.0 3 rd August 2017

2 PRIMIS: Quality Improvement Tool Instruction Guide Contents Introduction... 3 Aim of the diabetes care quality improvement tool... 7 Clinical audit notes and GP revalidation... 8 Running the diabetes quality improvement tool... 9 CHART Online... 9 Diabetes Case Finder View 1 Summary Sheet (classic view) Case finder summary Diabetes prevention programme summary View 2 - Datasheet Case finder pre-set filters Diabetes Care Management View 1 - Summary sheets Dashboard view Classic view NICE medication pathway blood glucose lowering therapy Diabetes medication safety summary View 2 Datasheet Diabetes care pre-set filters View 3 - CHART Online comparative analysis Key questions for GP practices Recommended follow-up work References Glossary Appendices Datasheet column list for diabetes case finder Datasheet column list for diabetes care Disclaimers How to create patient lists for use with mail merge functionality Diabetes_Care_V3.0 3 rd August 2017

3 PRIMIS: Quality Improvement Tool Instruction Guide Introduction Diabetes is a chronic metabolic disorder caused by defects in insulin secretion and action 1. There are two major types of diabetes, Type 1 and Type 2 along with a number of rarer types. Type 1 diabetes: The body is unable to produce any insulin because the insulinproducing cells of the pancreas have been destroyed 2. It can develop at any age but it typically develops in children or young adults (usually <40) 3. Treatment involves daily insulin injections, a healthy diet and regular physical activity. Approximately 10% of all adults with diabetes have Type 1. 2 Type 2 diabetes: Insulin is produced but either the amount is not sufficient for the body s needs, or the cells in the body do not react adequately to the insulin that is produced (insulin resistance). It is more commonly diagnosed in adults over 40 years, but increasing in younger people. Treatment involves a healthy diet, exercise, frequently oral medication and occasionally insulin is required 4. Approximately 90% of all people with diabetes have Type 2. 2 Conversely, around 95.1% of children (under 19) have Type 1 diabetes, whereas just 1.9% have Type 2. An additional 2.73% have Maturity-Onset Diabetes of the Young (MODY), Cystic Fibrosis related diabetes or their type is not defined 2. Approximately 1.1 million people in the UK have diabetes but have not yet been diagnosed 2. Five million people in England are at high risk of developing Type 2 diabetes with an HbA1c between 42-46mmol/mol. It is reported that the actual onset of diabetes can begin ten years or more before diagnosis and complications relating to diabetes may begin five to six years before diagnosis 2. Early diagnosis of diabetes, leading to the prompt provision of appropriate medical care, can save lives in people with Type 1 diabetes, and avert development or progression of complications of Type 2 diabetes, some of which can be lifethreatening. Action for diabetes, NHS England (2014) 5 Mortality in people with Type 1 diabetes is 131% greater than expected and 32% greater for those people with Type 2 diabetes. Both types are at increased cardiovascular risk 2. Data from the NDA for the last few years suggest that more than 24,000 people with diabetes die before their time each year in England and Wales. It is currently estimated that about 10 billion is spent by the NHS on diabetes. 10 per cent of the NHS budget The total cost (direct care and indirect costs) associated with diabetes in the UK currently stands at 23.7 billion and is predicted to rise to 39.8 billion by 2035/6. Facts and Stats, Diabetes UK (2016) 2 By the time they are diagnosed, half of the people with Type 2 diabetes are showing signs of complications. It is estimated that over 800,000 people with diabetes are at increased risk of complications due to poor glucose control 6. Diabetes_Care_V3.0 Page 3 of 68 3 rd August 2017

4 PRIMIS: Quality Improvement Tool Instruction Guide Recommended care processes Adults with Type 1 or Type 2 diabetes are largely managed in primary care however associated complications often result in hospital admission 1. The risk of complications associated with diabetes can be reduced by undertaking specific care processes as recommended in the Quality and Outcomes Framework (QOF), Action for Diabetes 5, the National Diabetes Audit (NDA), the State of the Nation 2016 report 17 and the NHS Outcomes Framework. There are 9 care processes that should be reviewed (at least) annually in all patients aged 12+ 7,8,9 : BMI measurement HbA1c measurement Record of smoking status Albumin: creatinine ratio Blood Pressure measurement Cholesterol measurement Foot examination Serum creatinine measurement Retinal screening The quality of NHS services for people with (or at risk of) diabetes has improved over recent years 6. In 2010 the UK had one of the lowest rates of early death due to diabetes compared to other wealthy countries 10. Data from the National Diabetes Audit (NDA) show a tenfold increase in the number of people who had received all nine diabetes care processes from 5% to 54% between 2005 and Data from the Quality and Outcomes Framework (QOF) for the individual care processes averaged 94% in 2009/10, which is significantly higher than that reported by the NDA* 11. It has been suggested that the emphasis on processes of diabetes care within the QOF and the NDA may be positively contributing to clinical outcomes such as reduced mortality rates 6. Taking action Whilst distinct progress is being made on improving clinical outcomes for people with diabetes in England, there is still room for improvement, particularly in relation to addressing geographical variations in standards of care 5. Data from the NDA in 2012/13 show big variations in the percentage of patients receiving the eight care processes typically delivered in general practice (the ninth, retinal screening, is provided externally). People living within the best performing CCG areas are 2.5 times more likely to receive eight care processes for diabetes than those living in the worst performing areas 2. There is also much work to be done in meeting the target levels in relation to HbA1c and blood pressure in addition to simply ensuring that the processes have been done. Opportunities for improvement also exist in relation to the prevention and early diagnosis of Type 2 diabetes and individual management of both Type 1 and Type 2 diabetes 5. Local Clinical Commissioning Groups (CCGs) and NHS England are under a statutory duty to continuously improve quality and NHS England has a statutory duty to conduct an annual assessment of every CCG. This is done using the CCG Improvement and Assessment Framework (CCG IAF) 12 which includes indicators relating to diabetes care. One indicator specifically looks at achievement of target values for three of the nine care processes listed above (HbA1c, blood pressure and cholesterol). Another indicator looks at the provision of structured education to newly diagnosed patients on how to manage their condition. Diabetes_Care_V3.0 Page 4 of 68 3 rd August 2017

5 PRIMIS: Quality Improvement Tool Instruction Guide *Data recording in general practice Some of the variation between the QOF and the NDA results can be explained by exclusions within the QOF for specific patients groups, such as newly registered patients or newly diagnosed patients and those aged under 17 years. Crucially though, the NDA and QOF were designed for different purposes; they utilise different code sets, different query logics and as a result the data are not directly comparable. 11 It is important to understand that the recording of data in general practice, using Read codes, is primarily influenced by the reason for recording that data electronically. The two main reasons (currently) are direct patient care and to achieve QOF target payments. As a result, there is often a strong incentive to use codes that are linked to QOF over other codes that may be available. Absence of a code within the patient s record does not necessarily mean that a particular care process has not taken place, but such information may instead be found in free text entries or elsewhere on the system (e.g. within scanned letters). This information still exists to support patient care but is not Read coded. 11 National programme support The PRIMIS diabetes quality improvement tool can help to support the delivery of a range of key national programmes of work in relation to diabetes. 1. Quality and Outcomes Framework Diabetes mellitus indicators: DM017, DM002, DM003, DM004, DM006, DM007, DM008, DM009, DM012, DM014, DM018. All of the data required for the current QOF indicators are included within the PRIMIS diabetes quality improvement tool. 2. NHS Diabetes prevention programme (NHS DPP) The NHS Diabetes Prevention Programme is a service developed via a partnership between NHS England, Public Health England (PHE) and Diabetes UK. It has been designed to target the five million people currently reported as being at high risk of developing type 2 diabetes. Between now and 2020, patients identified as being at risk will be referred onto a behaviour change programme funded by NHS England. 3. NICE Quality Standard [QS6] Diabetes in Adults 20 QS6 recommends that adults at high risk of type 2 diabetes are offered a referral to an intensive lifestyle-change programme (quality statement 1: preventing type 2 diabetes). Both NICE and the NHS DPP specify either a fasting plasma glucose level of mmol/litre or an HbA1c level of mmol/mol ( %) as an indication of high risk of type 2 diabetes. Diabetes_Care_V3.0 Page 5 of 68 3 rd August 2017

6 PRIMIS: Quality Improvement Tool Instruction Guide 4. NICE Clinical Guideline [CG181] Cardiovascular disease: risk assessment and reduction, including lipid modification 19 The PRIMIS diabetes tool incorporates the recommendations made within CG181 for assessment of cardiovascular risk in patients with diabetes. The care process section of the classic summary view (within CHART) highlights the number of patients satisfying the criteria specified within the NICE guidance as an alternative to achievement of an explicit Cholesterol level. 5. The NHS RightCare programme The PRIMIS diabetes quality improvement tool can help practices and CCGs to address a large number of the opportunity, spend and mortality indicators and clinical pathways specified within the RightCare programme. The relevant indicators and pathways are listed below: 1.2 Endocrine, Nutritional and Metabolic Disorders Opportunity Indicators Additional risk of complication among people with diabetes for... stroke [1.1, 1.2, 7.4] myocardial infarction [1.2, 7.4] heart failure [1.2, 7.4] The percentage of diabetic patients... whose last cholesterol was 5mmol or less [1.2, 1.5, 7.4] whose last HbA1c was 64 mmol/mol or less in last year [1.2, 7.4] whose last BP was 150/90 or less [1.2, 7.4] receiving all 8 care processes [1.2, 7.4] having retinal screening in last 12 months [1.2, 7.4] 2.2 Endocrine, Nutritional and Metabolic Disorders Spend Indicators Cost of endocrine - non-elective admissions [2.2, 7.4] Spend on primary care prescribing for endocrine [2.2, 7.4] 7.4 Diabetes Pathway Diabetes mellitus prevalence (17 years+) (%) Obesity prevalence (16+) (%) Spend on primary care prescribing for endocrine diabetes Cost of endocrine - diabetes - non-elective admissions The NHS RightCare programme has been designed to support the vision in the NHS Five Year Forward View by focusing on the transformation of healthcare services to drive improvements in quality and efficiency. Diabetes_Care_V3.0 Page 6 of 68 3 rd August 2017

7 PRIMIS: Quality Improvement Tool Instruction Guide Aim of the diabetes care quality improvement tool The aim of the diabetes care quality improvement tool is twofold; to assist with case finding activity and to report upon the level of care being offered to patients with known diabetes. This quality improvement tool expands the scope of the National Diabetes Audit in line with the NICE recommendations, as well as enabling practice/patient level analysis. The case finder element provides practices both with a list of patients identified as having possible diabetes and those at risk of developing diabetes (and therefore eligible for inclusion in the Diabetes Prevention Programme). By undertaking a review of these patients and adding any missing diagnosis codes, practices can improve the quality of their diabetes register, establish a more accurate prevalence rate and ensure that patients are monitored regularly and receive appropriate management. The care management part of the tool helps practices identify where they can improve the quality of care they provide to patients with known diabetes and reduce their risk of complications. The diabetes quality improvement tool enables practices to extract and analyse relevant coded data from their clinical information system. The tool works across all clinical information systems and presents data in an easy to use format allowing practices to gain insight and knowledge into their management of patients with diabetes. The diabetes quality improvement tool helps practices by: generating a list of patients with possible diabetes and providing relevant information to help clinicians to confirm or exclude diagnosis identifying patients at increased risk of developing type 2 diabetes and assessing eligibility for referral into the NHS Diabetes Prevention Programme allowing practices to achieve a more accurate prevalence rate for diabetes within their practice population identifying patients of all ages who have an existing diagnosis of diabetes and facilitating clinical audit against national standards establishing the number of patients with diabetes who have received the NICE recommended care processes in the last year and the number who have achieved the related target thresholds for HbA1c, BP and cholesterol reporting on the number of patients currently being treated in accordance with NICE recommendations regarding blood glucose lowering therapy and presenting data visually by populating the integrated NICE medication flowchart reporting on the number of patients with diabetes who are currently meeting NICE recommendations regarding CVD risk and lipid management reporting on the presence of associated complications of diabetes for all patients within the practice highlighting where prescribing in diabetes might be sub-optimal from a safety point of view with advice and cautions from the BNF Diabetes_Care_V3.0 Page 7 of 68 3 rd August 2017

8 PRIMIS: Quality Improvement Tool Instruction Guide identifying patients recently diagnosed with diabetes who have not been referred for structured education about their condition providing the facility to compare data with other practices both locally and nationally and the option to share aggregated data with their CCG via the CHART Online tool contributing to the delivery of the Quality and Outcomes Framework, National Diabetes Audit, NHS Outcomes Framework, Clinical Commissioning Group Improvement and Assessment Framework, NHS Diabetes Prevention Programme (NHS DPP) and the NHS RightCare programme. Please note that the focus of the quality improvement tool is adults with diabetes. Children with diabetes and women with gestational diabetes will appear in the results but they are not intended to be the main focus as they are largely cared for by secondary care services. PRIMIS quality improvement tools are designed simply to signpost GPs to patients who may be of interest or concern or may benefit from review. The tools are not intended to replace clinical decision making. Any action should be as a result of performing a clinical review with patients based upon individual circumstances. Clinical audit notes and GP revalidation This tool can be used to support GP revalidation. GPs can use the various CHART displays to review clinical data at both patient and practice level, enabling them to maintain an overall picture of how they manage patients at a population level but at the same time, looking in detail at the care of individual patients. This is a retrospective clinical audit. When conducting clinical audit for GP revalidation, GPs might choose to audit just their own clinical practice. Note that the diabetes tool will report on all patients with a diabetes diagnosis or with factors suggesting possible diabetes. Data will therefore be included on the activity of other colleagues. Involve fellow GPs in the clinical audit project. Several GPs who work together as a team can undertake a common audit. This is acceptable for the purpose of GP revalidation, as long as each GP can demonstrate they have contributed fully to the clinical audit activity. Alternatively, seek permission. A clinical audit on the care of patients with diabetes (or possible diabetes for case finder searches) matches the following criteria: it is of concern for patients/has the potential to improve patient outcomes it is important and is of interest to you and your colleagues it is of clinical concern it is of local or national importance it is practically viable there is new research evidence available on the topic it is supported by good research Diabetes_Care_V3.0 Page 8 of 68 3 rd August 2017

9 PRIMIS: Quality Improvement Tool Instruction Guide Running the diabetes quality improvement tool Before using the tool you must ensure that CHART is installed and you are familiar with how to use the software. Detailed instructions can be found on the PRIMIS website: There are two parts to the diabetes care quality improvement tool: one for the case finder and one for the management of patients with known diabetes. Within the CHART software, users can switch between the Diabetes case finder and the Diabetes report by using the Select Response Workbook function: Search inclusion criteria Searches are based upon patients who are currently registered at the practice. It is recommended that the data within the quality improvement tool are refreshed frequently (e.g. quarterly or six monthly) to monitor standards of care. CHART Online CHART Online is a secure web enabled tool that helps practices improve performance through comparative data analysis. By using CHART Online, practices can explore and compare the quality of their own data with anonymised data from other practices, locally or nationally, through interactive graphs. This provides a powerful tool for reducing variation across localities and may be of interest to local commissioning groups to facilitate the planning of care pathways. Aggregated summary data from both the diabetes case finder and diabetes care parts of the tool can be uploaded to CHART Online. Practices can view and download a personalised, detailed report that presents practice achievement alongside local CCG and national averages. CCG level comparative views are also available in addition to the practice specific reports. Please note that there is an inbuilt security function that prevents patient identifiable data being uploaded to CHART Online. Only aggregate data compiled from the pseudonymised responses can be transmitted. Diabetes_Care_V3.0 Page 9 of 68 3 rd August 2017

10 PRIMIS: Quality Improvement Tool Instruction Guide Diabetes Case Finder It is strongly recommended that practices use the case finder before going on to examine the management of patients with known diabetes. Using the case finder as a starting point will ensure that people with symptoms of diabetes are diagnosed earlier, receive appropriate treatment and that the practice diabetes register and practice prevalence rate are as accurate as possible. It will also ensure that those patients most at risk of developing type 2 diabetes are identified for referral to a structured prevention programme. The diabetes case finder consists of two separate summary sheets Case finder summary sheet Aimed at identifying patients who may have diabetes but do not have a coded diagnosis and patients at increased risk of developing diabetes Diabetes prevention programme (NHS DPP) summary sheet Aimed at identifying patients at high risk of developing type 2 diabetes and who are eligible for inclusion in the NHS DPP Both summary sheets help practices to answer the following questions: Do we have any patients with diabetes who do not have a diagnosis coded in their electronic record? How accurate is our practice prevalence rate for diabetes? Are there any patients who would benefit from a review for possible inclusion in the register and relevant treatment? Which patients are at greatest risk of developing the disease and require further glycaemic assessment? Which patients, considered at risk of developing type 2 diabetes, are eligible for referral to the NHS Diabetes Prevention Programme? How many patients have been assessed, referred, completed or declined inclusion in the NHS Diabetes Prevention Programme? View 1 Summary Sheet (classic view) The CHART summary sheets provide a synopsis of all the relevant data recorded by the practice. This is the best place to start when viewing the results. The diabetes case finder consists of two separate summary sheets. You can switch between them by clicking on the relevant hyperlink on the CHART summary sheet. Previews of the two summary sheets can be found on the following page. Diabetes_Care_V3.0 Page 10 of 68 3 rd August 2017

11 PRIMIS: Quality Improvement Tool Instruction Guide 1. Case finder summary sheet (preview) 2. Diabetes prevention programme summary sheet (preview) Diabetes_Care_V3.0 Page 11 of 68 3 rd August 2017

12 PRIMIS: Quality Improvement Tool Instruction Guide 1. Case finder summary Key information The first four rows of data (blue) provide some important pieces of information: an up to date count of the registered practice population the number of patients who have a recorded diagnosis of diabetes the number of patients with a diabetes resolved code (latest entry)* the recorded prevalence rate of diabetes in the practice What to note about this practice The recorded prevalence rate in this practice is 4.4% (for patients of all ages with diabetes). This is lower than reported national rates: o 6.5% from QOF 2015/16 15 (people aged 17+ registered on practice lists in England). *1 patient has been identified as having a diabetes resolved code. This patient may already have an existing coded diagnosis of diabetes but they also have a later entry of diabetes resolved (the resolved code is the latest entry). Patients can also appear in this section if they have a diabetes resolved code without a diagnosis. Suggested actions As a baseline quality check, assess whether the practice population count seems accurate. An unusually low number may suggest a problem whilst running the queries. Review the remaining summary sheet for further information on possible missing diagnoses and patients at risk. There may be patients with diabetes who have not yet had this coded which might explain any lower than expected prevalence rates in your practice. Reviewing the other data items highlighted by the case finder will help to find any patients with missing diagnoses. Diabetes_Care_V3.0 Page 12 of 68 3 rd August 2017

13 PRIMIS: Quality Improvement Tool Instruction Guide Table A Patients with a possible missing diagnosis code Table A is designed to identify patients with a missing or nonstandard diagnosis of diabetes, abnormal blood test or presence of monitoring codes which suggest the patient may have diabetes. The table gives the total number of patients identified along with a breakdown of the individual indicative factors. Patients may have more than one indicative factor therefore the count of total patients shown in the bottom row of the table may actually be lower than the sum of indicative factors that appear above. Tip: You can quickly access the filtered patient lists by clicking on the number cell. **To return to the summary view, click Show Summary on the ribbon/toolbar** Patients with a blood glucose level of 11.1 at any time are included for clinical relevance despite the fact that this diagnostic criterion does not form part of the WHO definition of diabetes. These patients may still be of interest. The results in Table A are based upon the datasheet column entitled possible missing diabetes diagnoses (see right). Within this column, patients can only be classified into one category. Where a patient has more than one factor, the one with highest precedence is displayed (regardless of date of entry). The precedence order is displayed in the table below from top to bottom, so diabetic medication takes precedence over non-standard diagnostic recording. Diabetic medication (last 12 months) Blood glucose diagnostic level (red) Latest 2 fasting blood sugar levels 7.0 Latest 2 HbA1cs 48 Diabetes monitoring code +ive Non-standard diagnostic recording As patients can have more than one factor, the counts provided by filtering on this datasheet column will not match the counts within the summary sheet table. Also, note that the presence of each of the factors in the filter drop down menu (see left) is entirely dependent upon at least one patient having data for that category. Diabetes_Care_V3.0 Page 13 of 68 3 rd August 2017

14 PRIMIS: Quality Improvement Tool Instruction Guide Important note: Patients in the diabetic medication category may be on diabetes related medication for other genuine reasons e.g. metformin for Polycystic Ovarian Syndrome (PCOS). Suggested actions If the total number of patients with a possible missing diagnosis is high, start by reviewing the most significant individual categories, for example, patients whose latest two HbA1c values were 48 (diagnostic range). Click on the relevant cell of the table to access the filtered list of patients. Use the associated data in the datasheet to gain a picture of the patient s history and look for clues as to whether a diagnosis may be missing or whether diabetes can be excluded. Some columns in the datasheet are hidden in order to present the most pertinent information first. In order to access hidden columns, click on the plus + signs as shown. The patient s full medical record may need to be examined and/or the patient called for review before a final diagnostic decision can be made. If, upon review, it is confirmed that the patient does not have diabetes but it is felt they may be at increased risk of developing the disease, they should be included in the NHS DPP. A risk assessment should be based upon recent glycaemic test results (ie. within the last 12 months), so these should be arranged if they have not been done. Consider using the mail merge function to produce invitation letters for patients that require a review. Patients in the yellow cell will appear in the first mail merge spreadsheet. Review this patient list before sending letters. For patients who have been assessed and diabetes excluded, consider adding Diabetes Mellitus Excluded to their electronic record (1I0.. in Read V2 or XaFvt in CTV3). This makes it clear that an assessment took place and when. These patients will need to be reviewed again in future, but the excluded column can be filtered to temporarily remove these patients from view if required. Diabetes_Care_V3.0 Page 14 of 68 3 rd August 2017

15 PRIMIS: Quality Improvement Tool Instruction Guide Table B - Patients at increased risk of developing diabetes Table B displays the number of patients at increased risk of developing diabetes and provides a summary of the contributory risk factors. Patients may have more than one contributory risk factor therefore the total number shown in the bottom row of the table may actually be lower than the sum of indicative factors that appear above. An additional column has been included to remove patients who have already appeared in Table A (these may already have diabetes but a coded diagnosis is absent). The purpose is to distinguish between those at risk and those who are likely to have missing diagnosis. Note: the category high glucose includes abnormal blood glucose test results or a diagnosis of impaired fasting glycaemia. Related datasheet column The Increased risk of diabetes column within the datasheet (shown right) allows you to filter the data by each of the individual risk factors. For the purposes of this filter, patients can only be classified into one category. Where a patient has more than one factor, the one with highest precedence is selected (regardless of date). The precedence order is displayed in the table below from top to bottom, so Polycystic Ovarian Syndrome (PCOS) takes precedence over Gestational diabetes. PCOS Glycosuria Insulin resistance High glucose Reaven's syndrome Prediabetes Impaired glucose tolerance Gestational diabetes Since patients can have more than one indicative factor, the counts provided by filtering on this column will not match the count of factors in the summary sheet table. Also, note that the presence of each of the factors in the filter drop down menu is entirely dependent upon at least one patient having data for that category. Diabetes_Care_V3.0 Page 15 of 68 3 rd August 2017

16 PRIMIS: Quality Improvement Tool Instruction Guide What to note about this example practice A total of 700 patients have been identified as being at increased risk of diabetes. 30 of these patients may already have diabetes but do not have a coded diagnosis (Table A patients). Once you exclude the patients who may have a missing diagnosis, there are 670 patients identified as being at increased risk of developing the disease (grey cell). Patients who appear in the grey cell may be eligible for the NHS DPP. Suggested actions You can quickly access the filtered patient lists by clicking on any number cell. To return to the summary, click Show Summary on the ribbon/toolbar. Click on the grey cell showing the number of patients at increased risk, excluding patients from Table A (670 patients in the example). These patients should be considered for the Diabetes Prevention Programme and are pulled through into Table E of the DPP assessment summary sheet (along with some patients from Table C). Use the associated data in the datasheet to gain a picture of the patient s history and look for indications or symptoms that may suggest diabetes has since developed. Establish when the patient was last reviewed and had a glycaemic assessment. The patient s full medical record may need to be examined and/or the patient called in for review. Refer to the section covering the DPP summary sheet for further advice on recommended actions. Patients identified as being at risk of diabetes should have had a recent glycaemic assessment (ie. within the last 12 months). Table E on the DPP summary sheet can help you identify these patients (see page 20). For patients who have been assessed and diabetes excluded, consider adding Diabetes Mellitus Excluded to their electronic record (1I0.. in Read V2 or XaFvt in CTV3). This makes it clear that an assessment took place and when. These patients will need to be reviewed again in future, but the excluded column can be filtered to temporarily remove these patients from view if required. Diabetes_Care_V3.0 Page 16 of 68 3 rd August 2017

17 PRIMIS: Quality Improvement Tool Instruction Guide Table C - Leicester Practice Risk Score* The case finder includes a Leicester Practice Risk Score calculator. This scoring system is recommended by NICE to identify: patients at risk of developing diabetes patients with non-diabetic hyperglycaemia who would be eligible for the NHS Diabetes Prevention Programme (NHS DPP) Please note this is a slightly different scoring system than that used in earlier versions of the case finder. This version is designed to calculate risk scores at practice level not individual patient level and is considered more accurate 18. The first column categorises patients into low or increased risk: Low risk <6 Increased risk 6 Data items used to calculate the patients scores include: age sex BMI ethnicity family history of diabetes anti-hypertensive use The second column extracted score shows the number of patients with a coded entry (38VZ. or XaeDt) of the Leicester diabetes score. This score would have been calculated elsewhere (e.g. within the practice) and the value entered onto the clinical system. Where an extracted score is present, this will take precedence over a calculation. The third column shows the numbers of patients in each risk category based upon a score calculated by CHART using all six data items. As all data items are present, the calculated score is considered accurate. The fourth column shows the numbers of patients in each risk category based upon a score calculated by CHART using partial data (ie. the patient does not have all six data items in their record). Patients identified as being at increased risk despite not having all data items present, can be reliably assumed to be at risk. Conversely, patients in the low risk category may not be truly low risk as the absence of all data items prevents calculation of a truly accurate score. The grey cell shows the total number of patients at increased risk of developing diabetes in the next ten years. This number does not include patients who appeared in table B and includes another set of patients who may be eligible for inclusion in the NHS DPP. These patients are pulled through into Table E of the DPP assessment summary sheet (along with some patients from Table B). *Developed by researchers within the Diabetes Research Centre at The University of Leicester 13 Diabetes_Care_V3.0 Page 17 of 68 3 rd August 2017

18 PRIMIS: Quality Improvement Tool Instruction Guide What to note about this example practice 410 patients have been identified as being at increased risk 32 patients have all six data items (used to calculate the score) present No patients have a coded entry of the Leicester diabetes risk score Suggested actions Click on the grey cell to access the full list of patients classified as being at increased risk of developing diabetes. These patients will appear in red on the datasheet. These patients should be considered for the Diabetes Prevention Programme and as a result they are included in the DPP assessment summary sheet Table E (see page 20). Use the associated data in the datasheet to gain a picture of the patient s history and look for indications or symptoms that may suggest diabetes has since developed. Establish when the patient was last reviewed and had a glycaemic assessment. The patient s full medical record may need to be examined and/or the patient called in for review. Refer to the section covering the DPP summary sheet for further advice on recommended actions. Compare the calculated risk score with the recorded/extracted score from the clinical system (if present see right). Review any discrepancies. Note that for the purposes of categorising patients by their risk score, CHART rounds the value to two decimal places within the datasheet. This can occasionally cause an artificial inflation/deflation in the datasheet column only (summary sheet is unaffected). Consider how data items used to calculate the Leicester risk score can be captured and recorded in order to increase accuracy of the calculated score. If key data items can be found elsewhere in the record (i.e. they have not been coded), ensure these are coded within the patient s record. Consider entering any scores calculated by CHART (based upon all six data items) onto the patient s record (38VZ. in Read v2 or XaeDt in CTV3). Also consider their QDiabetes risk score. Diabetes_Care_V3.0 Page 18 of 68 3 rd August 2017

19 PRIMIS: Quality Improvement Tool Instruction Guide Table D - Latest BMI value One of the factors of increased diabetes risk is a BMI 25. Table D gives a breakdown of BMI results 25 for all patients included in the datasheet. The table shows how many of these patients were also identified as possibly missing a diagnosis code (Table A), being at increased risk of developing diabetes (Table B) or with a Leicester risk score of 6 (Table C). What to note about this example practice 28 patients with a BMI 25 may also have a missing diagnosis 497 patients with a BMI 25 are also at increased risk of diabetes 372 patients with a BMI 25 also have a Leicester risk score of 6 Suggested actions Focus efforts on patients with a raised BMI who are also identified in Tables A, B or C. You may want to start with patients with the highest BMI. Access the relevant patient lists by clicking on the cells within the table. To return to the summary, click Show Summary on the ribbon/toolbar. Rounding issue in datasheet When categorising patients by their latest BMI, the value is extracted from the clinical system and then rounded to two decimal places within the datasheet. A subsequent column then identifies patients whose latest BMI 35. On occasion, this rounding up creates an artificial inflation, e.g. a patient s BMI of is rounded up to The Latest BMI 35 datasheet column will not include these patients as their true unrounded value is actually below 35 so they are correct to be excluded (see patient in the penultimate row in the example right). Diabetes_Care_V3.0 Page 19 of 68 3 rd August 2017

20 PRIMIS: Quality Improvement Tool Instruction Guide 2. Diabetes prevention programme summary Table E non-diabetic hyperglycaemia detection Table E pulls through patients identified in Tables B and C on the case finder summary. These patients were identified as being at increased risk of developing diabetes either by: the presence of indicative factors (Table B) or a Leicester diabetes risk score of 6 (Table C). What to note about this example practice In our example data, 670 patients were identified in Table B and 410 in Table C. Of these 1080 patients, 575 have not had a glycaemic assessment* in the last 12 months (see yellow cell in Table E). *The presence of either a HbA1c or a fasting blood sugar test counts in this category. In patients with Haemoglobinopathies a very small proportion of these may not be suitable for HbA1c use as a diagnostic /monitoring tool and fasting blood sugar values should be used instead (e.g. Rarer Thalassaemias and Homozygous Sickle cell disease). Most such patients are unaffected however. Suggested actions Access the filtered patient datasheet by clicking on the yellow cell. Use the associated data in the datasheet to gain a picture of the patient s history and look for indications or symptoms that may suggest diabetes has developed. Establish when the patient was last reviewed and had a glycaemic assessment. The patient s full medical record may need to be examined and/or the patient called in for review. Patients who have not had an HbA1c (or fasting blood sugar) in the last year should be tested. Pre-set Filter 4 lists patients deemed at risk (Leicester score) without recent glycaemic results. Consider using the mail merge function to produce invitation letters for patients requiring a review. Patients in the yellow cell appear in the second mail merge spreadsheet. Review this patient list before sending the letters. Diabetes_Care_V3.0 Page 20 of 68 3 rd August 2017

21 PRIMIS: Quality Improvement Tool Instruction Guide Table F Glycaemia assessment within the last 12 months For those patients who were identified as being at increased risk of developing diabetes (from Tables B and C) and who have had a recent glycaemic assessment (within the last 12 months), their test result will determine what action should be taken. Table F summarises this advice. What to note about this example practice In the example data 505 patients considered at risk had had a glycaemic assessment in the previous 12 months. Five out of the 505 patients have a raised HbA1c result (>47). 341 out of the 505 patients have a result that indicates non-diabetic hyperglycaemia and they are eligible for referral into the Diabetes Prevention Programme. 158 out of the 505 patients have a normal result. One patient had a glycaemic assessment but the value is missing. Suggested actions Option 1: If the patient s latest HbA1c value is >47 (or FBS >7), the test should be repeated (allow at least a two week gap before repeating the test). If the second result is also raised then the patient should be considered for a diabetes diagnosis. Due to the presence of two raised blood sugar results, these patients will now appear in Table A until a diagnosis code is added. Option 2: If the patient s latest HbA1c falls between (or FBS between ) they should be considered for referral into the NHS DPP. It is also recommended that these patients are coded with Non-diabetic hyperglycaemia ( C317. in Read v2 or XaaeP in CTV3). These patients will appear in Table G. Option 3: If the patient s latest HbA1c value is <42, the test should be repeated in 12 months. Missing values review the patient record to look for the missing test value. Diabetes_Care_V3.0 Page 21 of 68 3 rd August 2017

22 PRIMIS: Quality Improvement Tool Instruction Guide Table G Diabetes prevention programme status Table G gives a summary of diabetes prevention programme activity to allow practices to track progress and monitor activity. If patients have more than one NHS DPP code present, the latest entry will determine which category they fall into. The top row pulls patients through from Table F (grey cell) who were identified as being at risk based upon their latest glycaemic assessment (last 12 months). All of these patients should be coded with Non-diabetic hyperglycaemia as per NHS DPP recommendations. The CHART mail merge function can be used to produce invitation letters for patients eligible for the DPP with no previously recorded DPP codes (patients in the yellow cell will appear in the third mail merge spreadsheet). The middle row shows patients who already have an entry of Non-diabetic hyperglycaemia (ever). These patients may have already been referred into DPP and had a glycaemic assessment at the relevant time (ie. over 12 months ago). They appear here in order to track their progress through the programme. The bottom row shows NHS DPP activity for patients who do not have a coded entry of Non-diabetic hyperglycaemia. This allows you to view historic programme activity for patients who are no longer considered at risk because they have either become diabetic or reduced their risk level. The column top level programme code only provides information regarding the use of this generic code. It offers no insightful information, so practices should consider re-coding and using more specific codes for future activity. Diabetes_Care_V3.0 Page 22 of 68 3 rd August 2017

23 PRIMIS: Quality Improvement Tool Instruction Guide View 2 - Datasheet The datasheet is perhaps the most valuable part of the diabetes case finder. It allows practices to access the patient level data, providing all the relevant information in one place. The datasheet can be filtered as desired by the practice, to produce bespoke lists of patients. When preparing the queries to run on the clinical system, users are asked if they want to run a pseudonymised set, which uses a patient reference number (as shown below) or a patient identifiable set, that will return named patient information. The patient identifiable set is the most useful for case finding activity and also offers a mail merge function to allow practices to prepare invitation letters should they wish to call patients in for review. See the appendices for instructions. The CHART datasheet contains many columns of related data. A full list of available columns is included in the appendices of this document. For example, there is a column entitled Diabetes excluded date at the far right of the datasheet. This could be used to exclude patients who have recently had a diabetes diagnosis excluded (e.g. in the last six months). Apply a custom filter to the column by clicking on the grey drop down arrow, selecting date filters, after, then add the date from six or twelve months ago. Patients with an entry after this date will then be hidden when the filter is applied. Other useful columns within the case finder datasheet include: Latest NHS health check term this column will indicate whether the patient has been invited, attended or declined a health check. Latest Cancer Code this column gives practices the option of temporarily removing patients from the list who have a cancer diagnosis. These patients could then be manually removed from any mailing lists. Diabetes_Care_V3.0 Page 23 of 68 3 rd August 2017

24 PRIMIS: Quality Improvement Tool Instruction Guide Case finder pre-set filters In addition to creating custom filters, there are five pre-set (or pre-loaded) filters provided within the case finder tool. When viewing the datasheet, click Load Filter on the PRIMIS CHART toolbar. Load a filter as desired and then review the columns containing data items suggestive of diabetes to determine the value of reviewing the patients records in more detail. This will also assist with prioritising patients for review. To remove any filters, click clear on the CHART toolbar. Filter 4 aims to identify patients classified as at risk (based on Leicester risk scoring and age banding), who have not had a recent glycaemic assessment. These patients may be worth reviewing first. Filter 5 aims to identify patients classified as at risk (based on Leicester risk scoring and age banding), who have had a recent glycaemic assessment and are eligible for the NHS DPP based upon that result. Hidden datasheet columns Some columns in the datasheet are hidden in order to present the most pertinent information first. In order to access hidden columns, click on the plus signs towards the top of the datasheet (see image below): Once you have finished using the case finder tool and are confident about the accuracy of the practice diabetes disease register, move on to the next part of the quality improvement tool examining the care of patients with known diabetes. Diabetes_Care_V3.0 Page 24 of 68 3 rd August 2017

25 PRIMIS: Quality Improvement Tool Instruction Guide Diabetes Care Management It is recommended that practices use the case finder tool before going on to examine the management of patients with known diabetes. This will ensure that the practice diabetes register and prevalence rate are as accurate as possible before examining the care and management of patients with diabetes. The diabetes care management tool helps practices to answer the following questions: What is the practice prevalence rate for diabetes? How many patients have Type 1, Type 2, another type or have a diagnosis that does not define type? How many patients have received the recommended care processes in the last year? How many patients have achieved the relevant target thresholds for HbA1c, blood pressure and cholesterol? How many patients are currently being treated in accordance with NICE recommendations regarding blood glucose lowering therapy? How many of our patients with diabetes currently have complications associated with the disease? How many patients with diabetes are currently at risk due to sub-optimal prescribing? Diabetes care management output The diabetes care management tool provides the following views in CHART: 1. Summary sheets including - a dashboard view showcasing the practice s key results - a classic tabular view summarising the practice s key results - a medication safety summary - populated NICE blood glucose lowering/medication pathway 2. Full patient datasheet 3. A patient level datasheet specific to the eight care processes Detailed information on each of these data views can be found below. There is also a detailed individual practice report generated when data are uploaded to CHART Online. View 1 - Summary sheets CHART summary sheets provide a snapshot of all the relevant data recorded by the practice. For diabetes care management there are four different summary sheet views available as listed above. The dashboard view provides a visual display of the data whereas the classic view and medication summary present data in tabular form. Diabetes_Care_V3.0 Page 25 of 68 3 rd August 2017

26 PRIMIS: Quality Improvement Tool Instruction Guide Dashboard view Diabetes_Care_V3.0 Page 26 of 68 3 rd August 2017

27 PRIMIS: Quality Improvement Tool Instruction Guide Classic view The classic summary view contains some additional functionality and extra information that is not present on the dashboard, for example: cells with a double border contain active links to patient lists so users can quickly target patients requiring review who meet specific criteria. a summary of the numbers of newly diagnosed patients that have been offered or completed a structured education programme regarding their condition. other useful information such as the number of patients currently pregnant, using an insulin pump or with a read code indicating that they are at risk of unscheduled admission to hospital. Diabetes_Care_V3.0 Page 27 of 68 3 rd August 2017

28 PRIMIS: Quality Improvement Tool Instruction Guide Population/prevalence The classic view and dashboard view of the summary sheet start by providing key statistical information, including an up to date practice population count, disease prevalence rate and a breakdown of the number of patients with each type of diabetes. Untyped describes patients who have no specific diabetes type diagnosis code in their record (ie. a code that clarifies type). What to note about this example practice The practice prevalence rate for diabetes is 9.2%. This is higher than the 2015/16 Quality and Outcome Framework rate 15 of 6.5% for England. There are five patients who do not have a specific type of diabetes coded in their electronic health record. Suggested actions If your practice prevalence rate is inexplicably low compared to the national or local average (averages are included in the CHART Online reports) then consider looking for patients who are potentially missing a diabetes diagnosis. The case finder can help with this task. If your practice prevalence rate seems unusually high, review coding practice in this area or look for evidence of the underlying cause. Co-morbidities The second table on the classic view displays information about the number of patients with co-morbidities such as heart failure, AF, cardiovascular disease and hypertension. Patient lists can be quickly accessed by clicking on the cells with a double border. There is a corresponding graph on the dashboard. Knowledge of co-morbidities can help when planning a patient s care pathway particularly in relation to cardiovascular risk. It can also help you to understand how unwell patients are, or could become. It should be highlighted that patients picked up as having chronic kidney disease (CKD) may in fact have diabetic nephropathy but coding may not accurately reflect this therefore consider the rates for each. Diabetes_Care_V3.0 Page 28 of 68 3 rd August 2017

29 Complications Information regarding the incidence of complications relating to diabetes is important as they can often result in admission to hospital. Data can also help to inform appropriate management of the patient (e.g. visual impairment can make it difficult for patients to take the correct medication). It should be highlighted that patients with diabetic nephropathy may have instead been coded as having CKD. Consider rates for both CKD and diabetic nephropathy when reviewing data in this area. NOTES: The search criteria for the first six categories looks at entries ever. Diabetic ketoacidosis, foot ulcer and hypoglycaemia are only within the last 12 months. The denominator used to calculate the prevalence of erectile dysfunction is males with diabetes. Lower limb amputation excludes toe amputation only. Click the double border cells to access filtered patient lists. The first hypoglycaemia column in the table represents episodes where third party help may have been needed. The second column represents individuals who have information relating to minor hypoglycaemic episodes. High risk patients This summary gives information that is fundamental to the planning and provision of care for these patients. They may have difficulty with certain aspects of the self-management of diabetes. It includes information about mobility, learning disability, potential language barriers, carer status and whether they reside in a nursing or residential home. Diabetes_Analysis_V3.0 Page 29 of 68 3 rd August 2017

30 Processes of care The middle areas of both the dashboard and classic summary sheet display data regarding achievement of the recommended key processes of care for diabetes. As the provision of retinopathy services is largely outside the influence of the GP surgery, activity levels are reliant upon attendance rates being reported back to the practice. It is for this reason that retinal screening has not been included as one of the care processes in the central count section of the dashboard. Nonetheless, extracted information relating to retinal screening can be found within the main CHART datasheet. The top three lines of the classic view table (shown above) show whether the process/test has been done. To access lists of patients where the process or test has not been done, click on the red cells in the row Number NOT done. To return to this table, click Summary/Dashboard. Data are from the last 12 months with the exception of smoking status which is based on the QOF rules regarding codes and frequency of recording. The bottom three lines of the table relate to the achievement of any associated targets. The Achieving target row therefore shows the number of patients that have received that care process AND met the related target. In the example shown above, 96.4% of patients had their HbA1c measured but only 31% had values that met the target. Note that for patients under 12 years, there is only one recommended care process: HbA1c. Within the care process table two columns can be found that both relate to CVD risk/lipid management. The first relates to a standard cholesterol blood test with a general target of <5 mmol/l. This was set to <4 mmol/l in previous versions of this tool but is now aligned to national programmes such as NDA, QOF, NHS RightCare and CCG IAF. The second lipid column relates to the guidance specified within NICE CG181 for the assessment and management of CVD risk in patients with diabetes 19. The guidance differs by diabetic type and the tool will automatically assess patients according to this set criteria. By clicking on the blue hyperlink in the column title you can access a summary of the criteria used for this section. Diabetes_Analysis_V3.0 Page 30 of 68 3 rd August 2017

31 The centre of the dashboard gives a summary (image left) of the number of patients who have received all eight care processes (retinopathy screening is excluded). It also summarises how many of these patients are meeting recommended targets for HbA1c and blood pressure and whether patients at high risk of CVD have been prescribed a statin*. *or are contraindicated/declined. Criteria based upon NICE CG181 New (2017) functionality: Processes of care focused datasheet Thanks to development of CHART functionality PRIMIS are now able to create an array of relevant focused patient level datasheets. The diabetes tools contains a datasheet that specifically focuses on data relevant to the eight care processes. You can access this focused datasheet via the CHART toolbar. Navigation tip: When viewing the care process datasheet there are no + signs above the column titles and there are much fewer data columns available than when you are viewing the full datasheet (containing all data columns for the entire topic). To access the full datasheet from the care process datasheet, simply click DataSheet on the CHART menu. What to note about this example practice 61% of patients with diabetes received eight out of nine care processes in the last 12 months. 43.5% received all nine care processes. Only 6.9% of patients who received eight care processes had values for HbA1c and blood pressure that met the related targets and satisfied the NICE criteria for CVD risk management. 31% of patients with an HbA1c recorded in the last 12 months had values that were below (or on) target. This means that 69% of patients had an HbA1c above the target level. Suggested actions Review patients who are not meeting targets for HbA1c, blood pressure and lipid management and check that they are receiving optimum treatment. Review related tables and graphs from the quality improvement tool for more information on each target area (see pages of this guide). Review coding practice for any processes that are not well recorded (e.g. retinal screening). Investigate whether required information is being sent to the practice. Also, check that each process is routinely offered during annual reviews. Diabetes_Analysis_V3.0 Page 31 of 68 3 rd August 2017

32 Type 1 glycaemic control This dashboard graph displays information regarding glycaemic control in patients with Type 1 diabetes. This gives an indication of the risk of developing complications related to diabetes in the longer term. NICE targets 8 NICE guidance recommends setting an HbA1c target appropriate to the individual patient but that generally patients with Type 1 diabetes should be supported to achieve an HbA1c level of 48 mmol/mol (6.5%). Reassess individual needs and circumstances at each review. Lowering blood glucose levels reduces the risk of complications and high levels of HbA1c may indicate the need for glucose-lowering drugs. It is also recommended that the number of patients who achieve an HbA1c level of 53 mmol/mol (7%) or less should be documented. Graph key DCCT units IFCC units Category 6.5% 48 mmol/mol 6.6-7% % mmol/mol mmol/mol 10% 86+ mmol/mol Not done Within target Just above target - It is recommended that those with an HbA1c level of 53 mmol are documented. Above target - raised Above target - very high These patients are at increased hyperglycaemic risk which in turn is associated with increased risk of microvascular and macrovascular complications These patients are at significantly increased risk of complications such as heart failure 16 No HbA1c recorded in the last 12 months Suggested actions For patients above target, offer therapy (lifestyle and medication) to help achieve and maintain the HbA1c target level. Agree individualised HbA1c targets with patients taking into account the following factors: the person's daily activities, aspirations, likelihood of complications, comorbidities, occupation and history of hypoglycaemia. Check that patients are receiving optimal treatment as described within NICE guidance 8. Review patients with no HbA1c recorded in last 12 months. Diabetes_Analysis_V3.0 Page 32 of 68 3 rd August 2017

33 Non Type 1 Rx and HbA1c control This dashboard graph summarises patients HbA1c control and groups results by medication regime. NICE targets 9 Graph legend Green Within (or on) target Amber Above (or missed) target Red Grey Needs medication review Not done (no HbA1c recorded in the last 12 months) NICE recommends specific HbA1c targets according to treatment regime. DCCT units <6.5% 6.5% 7% IFCC units <48 mmol/mol 48 mmol/mol 53 mmol/mol Patients in this category Adults managed by lifestyle and diet Adults managed by lifestyle and diet combined with a single drug not associated with hypoglycaemia Adults on a drug associated with hypoglycaemia What to note about this example practice This practice appears to be good at monitoring patients HbA1c levels but may need to work further on appropriate management strategies. Suggested actions Review patients highlighted as needing a medication review (red) to ensure they are receiving optimal treatment as described within NICE guidance. Patients in the amber category who have missed their target could benefit from a review to establish any cause for unstable HbA1c. Offer support and diet and lifestyle advice as appropriate and monitor for potential progression into the red category. NICE guidance recommends setting a HbA1c target appropriate to the individual patient. Therefore, the targets specified here are provided as a guide (based on previous NICE guidance). Reassess individual needs and circumstances at each review. Diabetes_Analysis_V3.0 Page 33 of 68 3 rd August 2017

34 Blood Pressure (BP) control This graph displays information regarding achievement of blood pressure targets. No Rx indicates that patients are not currently on any antihypertensive medication. If BP is above target, then patients could be considered for this. Graph legend NICE targets 8 Green Within (or hit) target Target Patients in this category Amber Above (or missed) target <130/80 mm/hg Patients with eye, cerebrovascular or kidney damage (ECK) Red Not done (no BP recorded in the last 12 months) <140/80 mm/hg Patients without eye, cerebrovascular or kidney (ECK) damage What to note about this example practice There are some patients with eye, cerebrovascular or kidney damage (ECK) who have missed the target for blood pressure control in this practice. Suggested actions For patients whose blood pressure is above target, repeat measurement regularly in accordance with NICE guidelines. Give patients advice on lifestyle measures and check that they are receiving optimal treatment (as described within NICE guidance). Review patients with no blood pressure recorded in last 12 months. Diabetes_Analysis_V3.0 Page 34 of 68 3 rd August 2017

35 Lipid management and CVD risk This graph displays information regarding achievement of target cholesterol levels and categorises patients according to their use of a statin, alternative (non statin) therapy and those not using any lipid modifying therapy. Graph legend Green Amber Red Within (or hit) target Above (or missed) target Not done (no total cholesterol recorded in the last year) Patients with all types of diabetes are included in this graph. Note: Patients on statins may also be on other (non statin) lipid lowering therapy but they will only appear in the statin section of the above graph. Target cholesterol level vs NICE recommendations NICE no longer gives a specific target cholesterol level for people with diabetes. Instead it signposts clinical guideline CG181 which includes recommendations on lipid management in patients with diabetes 19. As a result, the diabetes care tool offers users an alternative way to assess progress towards the required care process for lipid management. Type 1 Non type 1 For the care process to be classed as For the care process to be classed as done and the associated target met, done and the associated target met, those patients with type 1 diabetes those patients with type 2 (or other who: non-type 1) diabetes who: are older than 40 years or have had diabetes for more than 10 years or have established nephropathy or have other CVD risk factors* are older than 85 years or whose latest QRISK2 10% or have other CVD risk factors* MUST ALSO HAVE been prescribed a statin in the last 6 months or a coded contraindication ever or a coded refusal ever *latest egfr <60 or CKD or Microalbuminuria or Familial Hypercholesterolaemia or IHD or CVD or Periperhal Arterial Disease MUST ALSO HAVE been prescribed a statin in the last 6 months or a coded contraindication ever or a coded refusal ever *latest egfr <60 or CKD stage 3-5 or Microalbuminuria or Familial Hypercholesterolaemia or IHD or CVD or Periperhal Arterial Disease Diabetes_Analysis_V3.0 Page 35 of 68 3 rd August 2017

36 NICE medication pathway blood glucose lowering therapy The following table provides information regarding treatment for patients with non type 1 diabetes only. This offers practices an overview of the numbers of patients in each treatment category. Patients will not be counted more than once. The integrated NICE blood glucose lowering pathway (see next page for full image) places patients with type 2 diabetes (only) into the relevant part of the pathway. In this view, patients will only appear once. Patients who appear in each red category are potentially not being treated optimally (in accordance with NICE guidance). For example, using the data shown to the left, there are 72 patients on diet alone who have an HbA1c above the target level of 48. These patients should be considered for commencement of Metformin. Of the 155 patients on metformin (with an HbA1c in the last year), 48 patients have a HbA1c above the target level of 58. These patients should be considered for dual therapy. To access the relevant lists of patients, click onto the box containing the number of patients you wish to see. To return to the pathway, click on the PRIMIS CHART toolbar at the top and then Show Summary. Diabetes_Analysis_V3.0 Page 36 of 68 3 rd August 2017

37 Diabetes_Analysis_V3.0 Page 37 of 68 3 rd August 2017

38 Diabetes education Other data items NICE recommend that all patients with diabetes are offered a structured education programme around the time of diagnosis 8,9. For Type 1 diabetes this should be 6-12 months after diagnosis or at any suitable point if not completed within the first year 8. The table above provides information on activity including refusals and DNAs. Click on the red cell to access the relevant patient list (those patients with no activity recorded). Number of patients pregnant in last year: This is provided for information and is a pointer to suggested audit. Pregnant patients with diabetes are cared for by (specialist) secondary care services and as a result primary care has little influence over their care during pregnancy. Number of patients with insulin pump: For information only. Admissions risk register: This section reports on the number of patients who have coded entries relating to the avoidance of unplanned admissions to hospital. The codes included in the search are as follows: Admission avoidance care started Admission avoidance care ended Admission avoidance care plan declined Read V2: 8CV4. CTV3: XaYD1 Read V2: 8CT2. CTV3: XaYD2 Read V2: 8IAe1 CTV3: XabFn CCG Improvement and Assessment Framework (IAF) Targets 12 This section gives summary information regarding achievement of the IAF targets relating to diabetes for HbA1c, blood pressure and cholesterol. This information is provided to facilitate any required reporting to CCGs by practices, should it be required. Diabetes_Analysis_V3.0 Page 38 of 68 3 rd August 2017

39 Diabetes medication safety summary The diabetes medication safety summary table helps practices examine areas where prescribing (in diabetes) might be sub-optimal from a safety point of view. There are two main groups; those of general prescribing safety and those related to poor renal function. Advice and cautions are taken from the BNF. Medication is analysed from the last six months. General prescribing safety Useful information Insulin and betablockers the number of patients identified is the number on insulin. The number of patients at risk is the number taking both insulin and betablockers. Women of child bearing age on statins - statins are contraindicated in pregnancy as a known cause of congenital abnormalities. This topic has been introduced to highlight the potential risk. Women of child bearing age on ACEI/ARB - ACE inhibitors and ARB drugs should be avoided in pregnancy unless essential. This topic has been introduced to highlight the potential risk. What to note about this practice There are 44 patients on both insulin and betablockers. There are 10 women of child bearing age on statins. There are 13 women of child bearing age on ACEI/ARB. Suggested actions Review the patients highlighted as being at risk. For women of child bearing age on statins or ACEI or ARB, ensure that they have been given appropriate advice about the risks in pregnancy. Diabetes_Analysis_V3.0 Page 39 of 68 3 rd August 2017

40 Renal function Useful information egfr data is based upon the latest entry in the last 12 months. Metformin and egfr - Increased risk of lactic acidosis means that metformin should not be prescribed at all if the egfr is less than 30ml/min, and should be prescribed with caution if the egfr is less than 45ml/min Sitagliptin and egfr - patients with an egfr<50ml/min should not take more than 50mg sitagliptin daily; note this does not take account of combinations with inherent metformin but they are handled elsewhere. Vildagliptin and egfr - caution is advised in patients on vildagliptin with an egfr<50ml/min; the specification above does include combination drugs including metformin which may be identified elsewhere if egfr<45ml/min. Saxagliptin and egfr - patients with an egfr<60ml/min should not take more than 2.5mg sixagliptin daily; the specification above does handle combination drugs including metformin on the assumption they are taken twice daily. Arcabose and egfr - Acarbose may accumulate in patients with renal impairment and should be administered with caution in patients with renal dysfunction. Exenatide and egfr - patients should avoid exenatide M/R where the egfr is <50ml/min and all exenatide forms where the egfr <30ml/min. Caution should be advised in patients on normal exenatide where their egfr <50ml/min. Diabetes_Analysis_V3.0 Page 40 of 68 3 rd August 2017

41 Liraglutide and egfr - patients should avoid liraglutide where the egfr <60ml/min. Alogliptin and egfr patients with an egfr <30ml should not take more than 6.25mg alogliptin daily and used with caution. Lixisenatide and egfr - patients should avoid taking Lixisenatide where the egfr <30ml/min, caution should be taken where the egfr is 30-50ml/min Dulaglutide and egfr - Dulaglutide is contraindicated where egfr <30ml/min Canaglifozin and egfr - for patients taking Canaglifozin the dose should be reduced to 100mg daily where the egfr <60ml/min persistently Dapaglifozin and egfr - patients should avoid taking Dapaglozin where their egfr <60ml/min as it is ineffective Empaglifozin and egfr - patients taking Empaglifozin should reduce their dose to 10mg daily if egfr <60ml/min, Patients should avoid taking Empaglifozin where their egfr <45ml/min What to note about this example practice Several patients have been identified as being at risk within this practice for example: o For example there are 28 patients on metformin with an egfr value of between 30 and 45ml/min Suggested actions Review the patients highlighted as being at risk and ensure they are being treated safely. Diabetes_Analysis_V3.0 Page 41 of 68 3 rd August 2017

42 View 2 Datasheet The datasheet (accessible via this icon from the toolbar) is perhaps the most valuable part of the diabetes care quality improvement tool. It allows practices to access patient level data, providing relevant information in one place to help clinicians review relevant information regarding diabetes care. The datasheet can be filtered as desired by the practice, to produce bespoke lists of patients. When preparing the queries to run on the clinical system, practices must decide whether to run a pseudonymised set, which uses a patient reference number (as shown below) or a patient identifiable set that will return named patient information. The patient identifiable set is the most useful for practices but to achieve the benefits of comparative analysis (using CHART Online) only the pseudonymised set can be uploaded in order to keep patient data secure. The CHART datasheet contains many columns of relevant data. A full list of available columns is included in the appendices of this document. As an example, towards the left hand side of the datasheet there is a column titled Latest diabetes type. It is beneficial for practices to apply a filter to this column to identify any patients who have an Untyped diagnosis (see image right). For these patients, there is no indication in their record as to whether they have Type 1, Type 2 or another type of diabetes which prevents patients being put on the correct course of treatment or audited effectively. It is likely that these patients have only ever had a high level code (eg. Diabetes Mellitus) added to their record. Diabetes_Analysis_V3.0 Page 42 of 68 3 rd August 2017

43 New (2017) processes of care datasheet The diabetes tools contains a datasheet that specifically focuses on data relevant to the eight care processes. Access this datasheet via the CHART toolbar. Navigation tip: When viewing the care process datasheet there are no + signs above the column titles and there are much fewer data columns available than when you are viewing the full datasheet (containing all data columns for the entire topic). To access the full datasheet from the care process datasheet, simply click DataSheet on the CHART menu. Diabetes care pre-set filters In addition to creating custom filters, there are four pre-set (or pre-loaded) filters provided within this quality improvement tool. Accessed via PRIMIS CHART, Load Filter when viewing the datasheet. Load a filter as desired and then review the columns containing data items to determine the value of reviewing the patients records in more detail. This will also assist with prioritising patients for review. Suggested action for practices For patients with a raised HbA1c, check they are receiving optimal treatment as described within NICE guidance. Review their individual HbA1c target to ensure it is appropriate. Review coding practice for annual reviews if numbers are low. Investigate whether checks are being done or whether annual review codes are not being added. Make efforts to review patients who have not had an annual review. Review patients who are on Sulphonylurea who do not have a contraindication to Metformin recorded. Add any missing contraindication codes. Check that patients are receiving optimal treatment as described within NICE guidance. Diabetes_Analysis_V3.0 Page 43 of 68 3 rd August 2017

44 View 3 - CHART Online comparative analysis By uploading aggregated summary data to CHART Online, practices can explore and compare the quality of their own data with anonymised data from other practices, locally or nationally. This provides a powerful tool for reducing variation across localities and can facilitate the planning of care pathways. Practices will be able to immediately view and download a personalised, detailed report that presents practice achievement alongside local CCG and national averages. Reports can be accessed through PRIMIS Hub. Diabetes_Analysis_V3.0 Page 44 of 68 3 rd August 2017

45 The individual practice CHART Online reports not only provide local and national comparators but also provide a narrative commentary on what practices need to do next. Graphs are provided for the following aspects of diabetes care: Key Performance Indicators 1. Number of care processes done 2. HbA1c target achievement Type 1 diabetes 3. HbA1c target achievement Non-Type 1 diabetes 4. Blood pressure targets 5. Cholesterol targets 6. Diabetes medication safety Diabetes_Analysis_V3.0 Page 45 of 68 3 rd August 2017

46 Key questions for GP practices Do we have any patients with diabetes who do not have the diagnosis coded in their electronic record? How accurate is our practice prevalence rate for diabetes? What should our approach be towards the number of patients identified with obesity and at increased risk of developing diabetes? How should we handle the broader group of patients at increased risk of developing diabetes? Which patients are eligible for inclusion in the Diabetes Prevention Programme? How can we improve our achievement of the NICE target thresholds for HbA1c and blood pressure? How effective is our approach to helping patients manage their glycaemic control? How effective is our approach to helping patients manage their cardiovascular risk? What is our strategy for reducing the risk of complications associated with the disease? Do we have a procedure in place to review patients whose treatment appears sub-optimal? Are key data items (such as the care processes) being recorded routinely and accurately? Should some of the individual patients identified be added to the practice Admissions Risk register? Recommended follow-up work Establish the proportion of the practice population at risk of developing diabetes. Consider and implement preventative strategies at both individual and population level. Review treatment efficacy with individual patients with known diabetes and consider alternative strategies as appropriate. Upload summative data to the PRIMIS CHART Online data warehouse and compare your practice data to other practices in the locality and nationally. Improve data recording and accuracy of clinical coding. Useful websites NHS Diabetes prevention programme NHS Right Care programme Diabetes_Analysis_V3.0 Page 46 of 68 3 rd August 2017

47 References 1. National Institute for Health and Care Excellence. (December 2013) Consultation on potential new indicators for the 2015/16 Quality and Outcomes Framework. Available: _Document_ pdf Last accessed: 6 th August Diabetes UK. (October 2016) Diabetes: Facts and Stats. Available: suk_facts_stats_oct16.pdf Last accessed: 3 rd August Diabetes UK. (March 2010) Diabetes in the UK 2010: Key statistics on diabetes. Available: Last accessed: 3 rd August Diabetes UK. (May 2014) What is Type 2 diabetes? Available: Last accessed: 3 rd August NHS England, Medical Directorate. (January 2014) Action for diabetes Available: Last accessed: 3 rd August Diabetes UK. (April 2012) Diabetes in the UK 2012: Key statistics on diabetes. Available: pdf Last accessed: 3 rd August Department of Health. (2001). National service framework for diabetes: standards. Available: Last accessed: 3 rd August 2017 Diabetes_Analysis_V3.0 Page 47 of 68 3 rd August 2017

48 8. National Institute for Health and Clinical Excellence. (August 2015) NICE Guideline 17: Type 1 diabetes in adults: diagnosis and management. Available: Last accessed: 3 rd August National Institute for Health and Clinical Excellence. (December 2015) NICE Guideline 28: Type 2 diabetes in adults: management Available: Last accessed: 3 rd August Murray, C.J.L., Richards, M.A., Newton, J.N., et al. (2013) UK health performance: findings of the Global Burden of Disease Study Lancet. Vol. 381(9871), pp Sullivan, R., and Easton, L (August 2012) Reporting on Annual Healthcare Checks for People with Diabetes. V12. Available: %20QOF%20Differences%20Report%20FINAL.pdf Last accessed: 3 rd August NHS England (May 2016) Technical annex for the CCG improvement and assessment framework 2016/17. Available: Last accessed: 3 rd August Gray, L.J., Taub, N.A., Khunti, K. et al. (2010) The Leicester Risk Assessment score for detecting undiagnosed Type 2 diabetes and impaired glucose regulation for use in a multiethnic UK setting. Diabet Med. Vol. 27(8), pp National Institute for Health and Clinical Excellence, (July 2012) NICE public health guidance 38. Preventing type 2 diabetes: risk identification and interventions for individuals at high risk. Available: Last accessed: 3 rd August NHS Digital (27 October 2016) Quality and Outcomes Framework Prevalence, Achievements and Exceptions Report, England Available: rep-v2.pdf Last accessed: 3 rd August Iribarren, C., Karter, A.J., Go, A.S. et al (2001) Glycemic control and heart failure among adult patients with diabetes. American Heart Association. Vol. 103, pp Diabetes_Analysis_V3.0 Page 48 of 68 3 rd August 2017

49 17. Diabetes UK. (2016) State of the Nation: Time to take control of diabetes. Available: s%20uk%20state%20of%20the%20nation% pdf Last accessed: 3 rd August Public Health England. National Cardiovascular Intelligence Network (NCVIN). (2015) NHS Diabetes Prevention Programme (NHS DPP) nondiabetic hyperglycaemia. Available: ta/file/456149/non_diabetic_hyperglycaemia.pdf Last accessed: 3 rd August National Institute for Health and Clinical Excellence. (September 2016) NICE Clinical Guideline 181: Cardiovascular disease: risk assessment and reduction, including lipid modification. Available: Last accessed: 3 rd August National Institute for Health and Clinical Excellence. (March 2011) NICE Quality Standard 6: Diabetes in Adults. Available: Last accessed: 3 rd August 2017 Diabetes_Analysis_V3.0 Page 49 of 68 3 rd August 2017

50 Glossary BMI BP CCG CCG IAF CHART CHART Online FBS HbA1c NDA NHS DPP NHS OF NICE PCT QOF QS Body Mass Index Blood Pressure Clinical Commissioning Group CCG Improvement and Assessment Framework PRIMIS data analysis tool (MS Excel based) PRIMIS comparative data analysis tool (web based) Fasting blood sugar Haemoglobin A1c. National Diabetes Audit NHS Diabetes Prevention Programme NHS Outcomes Framework National Institute for Health and Care Excellence Primary Care Trust (now replaced by CCGs) Quality and Outcomes Framework Quality Standard (NICE) Diabetes_Analysis_V3.0 Page 50 of 68 3 rd August 2017

51 Appendices 1. Datasheet column list for diabetes case finder Pseudonymised set Reference (MIQUEST pseudo ref) Both sets Age Sex Registered_Date Latest Ethnicity Latest BMI date Latest BMI value Latest BMI >=35 Penultimate HbA1c date Penultimate HbA1c value Latest HbA1c date Latest HbA1c value HbA1cs - same date Latest HbA1c in L12M Count of HbA1cs >=48 Penultimate fasting blood sugar date Penultimate fasting blood sugar value Latest fasting blood sugar date Latest fasting blood sugar value FBS - same date Latest fasting blood sugar in L12M Count of fast blood sugar >=7 Patient identifiable set Usual GP Reference (system ID number) NHS number Surname Forename Latest fast blood sugar >=5.5 - <7.0 Glycaemia assessmt in L12M expand + to see columns listed above Glycaemic assessment in DPP range GLYCAEMIC CONTROL SECTION Diabetes_Analysis_V3.0 Page 51 of 68 3 rd August 2017

52 Non-standard diagnostic recording Non-standard diagnostic recording date Diabetes monitoring code Diabetes monitoring date Diabetic medication L12M Diabetic medication L12M date Latest 2 HbA1cs >=48 Latest 2 fasting blood sugar >=7 Glucose >=11.1 date Glucose >=11.1 value Possible missing diabetes diagnoses expand + to see columns listed above Count of factors Gestational diabetes Gestational diabetes date PCOS PCOS date Prediabetes Prediabetes date Non-diabetic hyperglycaemia Non-diabetic hyperglycaemia date Prediabetes or non-diab hyperglycaemia Prediabetes or non-diab hyperglycaemia date Reaven's syndrome Reaven's syndrome date Insulin resistance Insulin resistance date Impaired glucose tolerance Impaired glucose tolerance date High glucose High glucose date Previous glycosuria diagnosis Previous glycosuria diagnosis date Coded at risk of diabetes 1408% Coded at risk of diabetes date Latest HbA1c >=42 and <48 Latest fasting blood sugar >=5.5 and <7.0 Increased risk of diabetes expand + to see columns listed above POSSIBLE MISSING DX INCREASED RISK DX Diabetes_Analysis_V3.0 Page 52 of 68 3 rd August 2017

53 Count of factors Hypertension diagnosis Hypertension diagnosis date Family history status Family history status date Extracted Leicester practice risk score date Extracted Leicester practice risk score No of Leicester practice risk score items Calculated Leicester practice risk score All Leicester practice risk score expand + to see columns listed above Risk score extracted or calculated Latest glucose date Latest glucose value Glucose tolerance test Glucose tolerance test Date Glycosuria Glycosuria date Triglycerides value date Triglycerides value Latest HDL date Latest HDL value Latest smoking QOF expand + to see blood test columns listed above Latest smoking QOF date Current smoker code Current smoker date expand + to see smoking columns listed above Latest FH code or 1st degree Latest FH code or 1st degree date Qdiabetes score date Qdiabetes score High diabetes risk High diabetes risk Date IHD IHD Date Latest BP medication Latest BP medication date Cushings syndrome Cushings syndrome date Oral steroids rx INCREASED RISK DX OTHER BLOOD TESTS SMOKING STATUS OTHER RISK FACTORS Diabetes_Analysis_V3.0 Page 53 of 68 3 rd August 2017

54 Oral steroids rx date Raised HbA1c >=42 ever date Raised HbA1c >=42 ever value Raised fasting blood sugar >=5.5 ever date Raised fasting blood sugar >=5.5 ever value Waist circumference date Waist circumference value Latest BMI>=25 ever date Latest BMI>=25 ever value Latest cancer code Latest cancer term Latest cancer date expand + to see risk factor columns listed above Eligible for DPP NHS DPP programme code NHS DPP programme date NHS DPP referred code NHS DPP referred date NHS DPP declined code NHS DPP declined date NHS DPP started code NHS DPP started date NHS DPP not completed code NHS DPP not completed date NHS DPP completed code NHS DPP completed date Latest DPP code Latest DPP term Latest DPP date expand + to see columns listed above Latest NHS health check code Latest NHS health check term Latest NHS health check date expand + to see columns listed above Diabetic diagnosis date Diabetes resolved date Diabetes excluded date expand + to see columns listed above OTHER RISK FACTORS DPP PROGRAMME RECORDING HEALTH CHECKS OTHER DIAB DX Patient group Need HbA1c or FBS Diabetes_Analysis_V3.0 Page 54 of 68 3 rd August 2017

55 2. Datasheet column list for diabetes care Pseudonymised set Reference (MIQUEST pseudo ref) Both sets Age Sex Registered_Date Earliest Diabetes Diagnosis Code Earliest Diabetes Diagnosis Date Latest Diabetes Type 1 Code Latest Diabetes Type 1 Date Latest Diabetes Type 2 Code Latest Diabetes Type 2 Date Latest Probable Type 1 Code Latest Probable Type 1 Date Latest Probable Type 2 Code Latest Probable Type 2 Date Latest Other Diabetes Code Latest Other Diabetes Date Latest Untyped Diabetes Code Latest Untyped Diabetes Date Patient identifiable set Usual GP Reference (system ID number) NHS number Surname Forename Latest Diabetes Type expand + to see columns listed above Ethnicity code Housebound code Housebound date Long stay code Long stay date Carer status code Carer status date Vulnerable elderly date On learning disability register date Poor English code Poor English date High risk care expand + to see columns listed above DIABETES DIAGNOSIS ETHNICITY & HIGH RISK CARE Diabetes_Analysis_V3.0 Page 55 of 68 3 rd August 2017

56 Albumin creatinine ratio date Albumin creatinine ratio value egfr code egfr date egfr value Creatinine date Creatinine value Total cholesterol date Total cholesterol value Total cholesterol level HbA1c target value HbA1c target date HbA1c date HbA1c value expand + to see blood test columns listed above Latest weight date Latest weight value Latest BMI date Latest BMI value Latest BMI level expand + to see columns listed above Blood pressure date Blood pressure systolic Blood pressure diastolic ECK BP level expand + to see blood pressure columns listed above Latest foot ulcer status Amputation left code Amputation left date Amputation right code Amputation right date Left foot risk code Left foot risk date Right foot risk code Right foot risk date Foot check status expand + to see foot screening columns listed above BLOOD TESTS WEIGHT BLOOD PRESSURE FOOT SCREENING Diabetes_Analysis_V3.0 Page 56 of 68 3 rd August 2017

57 QOF smoking code QOF smoking date Smoking brief intervention date Smoking cessation referral code Smoking cessation referral date Smoking cessation activity code Smoking cessation activity date Smoking cessation medication code Smoking cessation medication date Smoking status expand + to see smoking status columns listed above Count of care processes Retinal screening date Diabetic neuropathy date Diabetic retinopathy date Diabetic nephropathy date Visual impairment registration code Visual impairment registration date Amputation code Amputation date Diabetic ketoacidosis date Erectile dysfunction date Hypoglycaemia L12M code Hypoglycaemia L12M date Minor hypoglycaemia date Count of complications expand + to see complication columns listed above Earliest CKD diagnosis code Earliest CKD diagnosis date Latest CKD stage code Latest CKD stage date Latest CKD diag or term Earliest IHD diagnosis date Stroke or TIA date Earliest hypertension diagnosis date Earliest heart failure diagnosis date Earliest AF diagnosis date Count of co-morbidities Earliest dementia diagnosis date QRisk value QRisk date SMOKING STATUS COMPLICATIONS CO-MORBIDITIES Diabetes_Analysis_V3.0 Page 57 of 68 3 rd August 2017

58 FHx Hypercholesterolaemia code FHx Hypercholesterolaemia date Albuminuria code Albuminuria date PAD code PAD date expand + to see comorbs & QRisk columns listed above Metformin Rx code Metformin Rx date Adverse reaction to metformin code Adverse reaction to metformin date Metformin Rx with no later CI Sulphonylurea Rx code Sulphonylurea Rx date Metformin & su Adverse reaction to sulphonylurea code Adverse reaction to sulphonylurea date Glibenclamide Rx code Glibenclamide Rx date Adverse reaction to glibenclamide code Adverse reaction to glibenclamide date Thiazolidinedione Rx date Adverse reaction to thiazolidinedione code Adverse reaction to thiazolidinedione date Bladder cancer diagnosis date DPP4 inhibitor Rx code DPP4 inhibitor Rx date DPP4 inhibitor Rx term SU & DPP4 GLP-1 stimulator Rx code GLP-1 stimulator Rx date GLP-1 stimulator Rx term Meglitinide Rx code Meglitinide Rx date Insulin Rx code Insulin Rx date Adverse reaction to insulin code Adverse reaction to insulin date Metformin s-urea & insulin Acarbose Rx date SGLT-2 Rx code COUNT OF MEDICATIONS Diabetes_Analysis_V3.0 Page 58 of 68 3 rd August 2017

59 SGLT-2 Rx date Count of medications expand + to see diabetic Rx columns listed above Statin Rx code Statin Rx date Statin declined or contraindicated (code) (Statin) declined or contraindicated date Non statin cholesterol lowering Rx (code) Non statin cholesterol lowering Rx date ACEI Rx code ACEI Rx date ACEI CI or declined code ACEI CI or declined date ARB Rx code ARB Rx date ARB CI or declined code ARB CI or declined date ACEI or ARB Beta blocker date Hypertension Rx code Hypertension Rx date Maximum tolerated antidiabetic Rx expand + to see columns listed above Seasonal flu vaccination date Seasonal flu vaccine date Seasonal flu vacc declined date Flu vac done or declined Pneumococcal vaccination date Pneumococcal vaccine date Pneumococcal declined date Pneumo vacc or declined expand + to see columns listed above Depression diagnosis code L12M Depression diagnosis date L12M Depression resolved date Depression screening questions code Depression screening questions date Latest depression status expand + to see columns listed above CHOLESTEROL & HYPERTENSION RX VACCINATIONS DEPRESSION DX & SCREEN Diabetes_Analysis_V3.0 Page 59 of 68 3 rd August 2017

60 Dietary review date Dietary review declined date Earliest structured education (code) Earliest structured education term Earliest structured education date Diagnosed 12 to 24 months ago Structured education <12mts after earliest Dx Latest structured education (code) Latest structured education term Latest structured education date Structured education status Erectile dysfunction screening date Erectile dysfunction advice date Latest microalbiminuria date Latest persistent proteinuria date Dementia screening questions code Dementia screening questions date Diabetes care plan declined date Diabetes care plan date Diabetes self-management plan code Diabetes self-management plan date Insulin pump date Pregnant in last year code Pregnant in last year date Hysterectomy date Female sterilisation Coil in place code Coil in place date Coil removal date Diabetic care hospital only date Secondary care involved code Secondary care involved date Admission risk register code Admission risk register date Diabetes annual review date expand + to see columns listed above SUPPLEMENTARY INFORMATION Diabetes_Analysis_V3.0 Page 60 of 68 3 rd August 2017

61 HbA1c category First intensification Second intensification Metformin CI 1st step SU/DPP4/TZD Metformin CI 1st step DPP4/TZD Metformin CI 1st intensification Metformin CI 2nd intensification Care process - smoking BP target Care process - BP Cholesterol target Care process - cholesterol NICE cholesterol risk NICE cholesterol care process NICE cholesterol target Care process ACR HbA1c Target Care process - HbA1c Care process - BMI Care process - Serum Creatinine Care process - foot check ECK & BP meds Cholesterol Rx for graph Women of child bearing age expand + to see columns listed above GRAPH WORKINGS Diabetes_Analysis_V3.0 Page 61 of 68 3 rd August 2017

62 3. Disclaimers Risk scoring Calculation of the Leicester practice risk score for diabetes is dependent upon certain codes being present within the patient s electronic record. Absence of these codes could either indicate that the patient does not have the specified risk factor or that alternative Read codes could be being used that are inaccurate or too generic. It is pertinent therefore that practices record such clinical data in as much detail as is possible and is relevant. Scores provided within this quality improvement tool should not replace clinical decision making and are only included to help inform that decision. Patients must be reviewed to confirm the accuracy of recorded information before management is decided. Mail merge patient lists Although every care has been taken to ensure the accuracy, completeness and reliability of the diabetes case finder tool, we advise that a health care professional validate the output of the mail merge facility prior to any letters being sent to patients. Accordingly, and in line with our terms and conditions of membership (section 10), PRIMIS, a business unit of The University Of Nottingham, disclaims any and all liability resulting from, or related to, performance issues associated with use of its diabetes case finder tool. Diabetes_Analysis_V3.0 Page 62 of 68 3 rd August 2017

63 4. How to create patient lists for use with mail merge functionality 1. When you import the named response files into CHART, you will be given the option to create a mail merge file 2. If you select Yes, you will then be asked which file you want to create. There is a choice of three different files to create or you can create them all (option 4). Type the option number you require then click OK. A. Possible missing diabetes diagnosis this includes patients with read coded entries suggestive of diabetes, but who do not have a (standard) coded diagnosis. They are the same patients that appear in the yellow cell of Table A within the case finder summary sheet (see below). Diabetes_Analysis_V3.0 Page 63 of 68 3 rd August 2017

64 PRIMIS: Quality Improvement Tool Instructional Guide B. Patients at increased risk of diabetes with no glycaemic assessment in the last 12 months this includes patients from tables B and C from the case finder summary sheet who are considered at risk of developing diabetes, but have not had a recent glycaemic assessment (last 12 months). This list can be used to call these patients in for a blood test. They are the same patients that appear in the yellow cell of Table E within the case finder DPP summary sheet (see below). Patients who were included in Table A or mail merge list 1 are excluded here. C. Patients whose latest HbA1c is between 42 and 47 or fasting blood sugar is between 5.5 and 6.9 with no Diabetes Prevention Programme (NHS DPP) code Patients with glycaemic results within these ranges are eligible for referral into the NHS DPP. This list can be used to include patients in this programme. They are the same patients that appear in the yellow cell of Table G within the case finder DPP summary sheet (see below). Patients who were included in mail merge list 1 (possible missing diabetes diagnosis) or mail merge list 2 (at risk who require glycaemic assessment) should not appear in this list (at risk based upon recent glycaemic assessment but no NHS DPP activity). Diabetes_Analysis_V3.0 Page 64 of 68 3rd August 2017

65 PRIMIS: Quality Improvement Tool Instructional Guide 3. The following message box will appear before you see the CHART Summary sheet: 4. Upon clicking OK, you will see the following message box. Please browse to choose a location to save the files (creating a new folder if required). As these files contained patient identifiable data, you should store them in a secure location. 5. Now open your own created template letter. If you haven t got a standard practice template letter, you must create this first and add the relevant merge fields available from the lists. Simply create a letter in MS Word and then click on the Mailings tab. Click Start Mail Merge, Letters. 6. Type out the content of your letter. Diabetes_Analysis_V3.0 Page 65 of 68 3rd August 2017

66 PRIMIS: Quality Improvement Tool Instructional Guide 7. Choose Mailings from the ribbon, then click on Select Recipients and Use Existing List. 8. Browse to the location of the mail merge list stored earlier and click Open : 9. Click on OK when you see the following message: 10. Click Finish & Merge and Edit Individual Documents. 11. Click OK to merge to a new document. Diabetes_Analysis_V3.0 Page 66 of 68 3rd August 2017

67 PRIMIS: Quality Improvement Tool Instructional Guide 12. You will then be able to scroll through the letters, print them or browse to a location to save them for printing at a later date. It is strongly recommended that a clinician reviews the list of patients provided before any mail merge letters are created and sent out. 13. Repeat steps 1-9 for each of the three lists of patients. Word 2003: 1. After opening and editing the letter, choose Letters and Mailings from the Tools menu, then click on the option Show Mail Merge Toolbar. Diabetes_Analysis_V3.0 Page 67 of 68 3rd August 2017

68 PRIMIS: Quality Improvement Tool Instructional Guide 2. Click on the Open Data Source button on the toolbar. 3. Browse to the location of the relevant mail merge list, select it and click Open. 4. Click Ok when you see the following message box: 5. Go back to Word and click on the View Merged Data button on the toolbar. 6. You will then be able to scroll through the letters, print them or browse to a location to save them for printing at a later date. It is strongly recommended that a clinician reviews the list of patients provided before any mail merge letters are created and sent out. Diabetes_Analysis_V3.0 Page 68 of 68 3rd August 2017