Compiling and Maintaining a Diabetes Register in Primary Care  

Contents

• Introduction: the importance of registers
  - Definition of Register
  - Required Properties of Register
  - Aspects of Register
  
• Compiling an up to date Diabetes Register
  
• Maintaining the Register
  - Table 1: guidance on managing registers
  - Clinical Audit
  - Templates, including key points to consider
  
• Appendix 1: Diagnostic Criteria for Diabetes
  - Methods and Criteria for Diagnosing Diabetes Mellitus
  - Classification and Terminology
  
• Appendix 2: Evaluation of Registers
  
• Appendix 3:  Electronic Register
  
• Appendix 4: Query about Computerised Diabetes Register

Introduction – the importance of registers

Definition of a register

A register is a realisable accurate list of patients who share a common characteristic usually a disease entity.

This list need not exist until it is needed. Indeed this is desirable because the dynamic nature of primary care means that any list is out of date very soon after its creation. What is important is that the codes used to generate the list are kept up to date. The list system in primary care ensures that patients are automatically removed from the register when they die or move away.   

Required Properties of a register

Any method of register construction must have the following properties:

1. Easily maintained
2. Applicable to a wide range of possible conditions
3. Accurately distinguish between patients with the condition and without it, i.e. it must not pick up patients who haven’t got the condition or miss those with it.
4. Is easy to keep up to date.

Aspects of Registers 

Creation:

      Disease definition
Define dual and mutually exclusive coding (see below)
Patient identification
Procedure for data capture – paper flows and templates
Coding of patients – systematically add from list using templates 

Maintenance:

      Procedures for handling incoming information and data entry.
Procedures for handling  existing data e.g.. notes summarising.
Staff training and responsibilities for data input. 

Housekeeping:

      Dealing with false negatives i.e. when patient is not identified who should be.
Dealing with false positives i.e. when patient is registered who shouldn’t be.

Compiling an up to date Diabetes Register 

An up to date register can be compiled using the following steps

1) Identify patients with the diagnosis of Diabetes Mellitus from the computer using searches on the following Read Codes. 

For a patient to be counted as on the register they need: 

C10 – Diabetes Mellitus Code to be present.
AND EITHER
C10E           Type I Diabetes
C10F           Type II Diabetes

If more than one of each code exists this does not present a problem, BUT the earliest C10F/C10E should be the date of diagnosis. This is important because the audit criteria specify that certain eye and renal checks do not need to be performed in Type I diabetics until five years after diagnosis. 

2) Add to the list additional names generated by searching for repeat prescription drugs used for Diabetes Therapy.

3) Further names can be added by health care professionals in the PHCT (circulate and add to the list generated from 1 and 2) from memory/other records e.g. district nurse records. 

4) A list from the health authority of hospital activity data can provide a list of patients who have been seen in secondary care. The Primary Care Trust will need to organise this. 

5.             Steps 1 – 3 will identify the majority of established diabetics in the practice.  However the register can be further updated by adding patients as they consult opportunistically or as hospital letters are received.  This requires a manual or electronic prompt on the record as to whether or not a patient is already on the register.  (Stickers on notes can be used for example). 

6.             For computerised practices, steps 1 and 2 are automatically carried out by using the Leicestershire PCAG housekeeping set (see electronic audit pages) . This generates the appropriate lists for checking along with diabetes diagnoses and is quicker than setting up all the searches individually.

Maintaining the Register  

1. Computerised practices will need to ensure that the appropriate Read Code data is added to the system as new patients are detected or join the practice.  (Similarly patients who leave must be removed from the active register)

2. Practices with manual systems will need to add and delete patients from a manual system as necessary.

Both 1 and 2 require robust administrative systems and a protocol in the practice which makes it clear who is responsible for firstly identifying the relevant patient and secondly recording on the appropriate database.  It is recommended that a named administrative lead is identified.

For practices using electronic registers the housekeeping set can be used to maintain the register. This can be done by intermittently running the Leicestershire PCAG housekeeping set particularly two weeks before an audit is carried out. It also provides a useful check on whether the practice procedures for handling data are working as well as helping to identify where the system might be falling down.

Table 1 provides guidance on how to manage paper based or electronic registers 

Clinical Audit   

Clinical Audit allows an opportunity to update the register.  Audit should include review of the diagnosis and whether the patient is still registered with the practice.  Audit data will also provide a practice with prevalence data, which may indicate if numbers of patients on the register are accurate for a given population. Clinical audit will also provide data on process and outcome measures being achieved for people with diabetes.

Templates  

Computerised practices, which use templates (see electronic audit pages), will not need to maintain a special separate register.  The general computer database will act as the Register Database with the template providing the basic audit information.  If recorded correctly this can easily and quickly provide required audit data including appointments due.

Templates and Registers – Key Points to consider

Diabetes Computer Templates – Practices need to:

•  Be familiar with the template itself (there are variations in the available templates, see PCAG  Website for details).
  

• Know the purpose of the template. It is a simple and useful method of recording diabetic data. It requires accurate information to be entered and updated. If this is routinely done it will prompt decision-making for professionals regarding treatment and/or management.
  

• Collect minimum data set as a first step.
  

• Accurate and comprehensive information collected - better prompts to the professional.
  

• Internal monitoring will promote advanced entry over time.

Searching – Practices need to identify individuals capable of undertaking the following:

• Search using EMIS or other system VAMP VISION, TOREX etc.
  

• Search by Read code, by name, by condition etc.
  

• Be comfortable using the system owned by the practice.
  

• Recognise agreed Diabetic Read Codes (C10 and either C10E or C10F)
  

• For EMIS practices using the diabetes template entering the diagnosis will automatically Read – Code the entered data.
  

• Accept that there must be an access password imposed to protect staff and patients. This means that only certain individuals can access certain information which will be controlled by allocation of access pass-wording or similar safety and risk management features – e.g. GPs (or designated other) only able to prescribe or alter medications.  GPs (or designated other) only able to change diagnoses.
  

• If PRIMIS is being used in the practice any training should be linked/ integrated with PRIMIS by coordinators as this will ensure data quality.

APPENDIX 1

Diagnostic Criteria for Diabetes (WHO Classification) 

Summary of Changes

The World Health Organisation has published revised ‘Definition, Diagnosis and Classification of Diabetes Mellitus and its Complications’, and Diabetes UK has recommended that all healthcare professionals adopt the new criteria from 1 June 2000.  The main changes of the new recommendations are set out below.  They include the recommendation that the cut off point for diagnosing diabetes using a fasting plasma glucose should be lowered from 7.8 mmol/l.  This change reflects research evidence regarding the development of complications of diabetes.

Methods and Criteria for Diagnosing Diabetes Mellitus  

1. Diabetes symptoms (i.e. polyuria, polydipsia and unexplained weight loss) plus a      random venous plasma glucose concentration > 11.1mmol/l or a fasting plasma           glucose concentration > 7.0mmol/l.  (Whole blood > 6.1 mmol/l) or 2 hour plasma        glucose concentration > 11.1mmol/l 2 hours after 75g anhydrous glucose in an oral           glucose tolerance test (modified OGTT).
   

2. With no symptoms diagnosis should not be based on a single glucose determination but requires confirmatory plasma venous determination.  At least one additional glucose test result on another day with a value in the diabetes range is essential, either fasting, from a random sample or from the two hour post glucose load.  If the fasting or random values are not diagnostic the modified GTT should be used.

Classification and Terminology 

The terms Type 1 and Type 2 have replaced IDDM and NIDDM, and Type 1 and Type 2 process has been introduced to describe the cause of insulin-dependent and non-insulin dependent diabetes respectively. 

Impaired Glucose Tolerance (IGT)* is a stage if impaired glucose regulation (fasting plasma glucose < 7.0mmol/ and OGTT 2-hour value > 7.0mmol/l but < 11.1mmol/l. 

Impaired Fasting Glycaemia (IFG)* has been introduced to classify individuals who have fasting glucose values above the normal range but below those diagnostic of diabetes.  (Fasting plasma glucose > 6.1mmol/l but < 7.0mmol/l) 

Some concern has been expressed about the implications of these changes for diabetes care.  The new criteria have clarified the diagnosis of diabetes and the ability to diagnose cardiovascular high risk cases in many people.  Earlier diagnosis will increase the total number of people with diabetes, but if they are managed according to Diabetes UK guidelines, many of these new cases will be diet controlled.  In the long term, complications should be reduced,  to the benefit of the individual and to the health service. 

*IGT and IFG are not clinical entities in their own right but rather risk categories for cardiovascular disease (IGT) and/or future diabetes (IFG) 

The full text of the revised diagnostic criteria for diabetes can be found on the Diabetes UK website (www.diabetes.org.uk)

APPENDIX 2 

Evaluation of Electronic Registers 

Disease registers should contain as few patients falsely identified as having the disease (false positives) as possible. There should also be as few patients not on the register who actually have the disease (false negatives) as possible. 

The Housekeeping procedures devised by the Leicestershire Primary Care Audit Group which use Miquest searches which exploit the dual coding design of disease registers, enable checks to be carried out on disease registers to identify both these groups and to amend the electronic record accordingly using the template. It is relatively easy to assess the completeness of a register (i.e. the absence of false negatives).  By comparing the practice’s disease prevalence, as estimated by the disease register and practice list size, with an expected practice prevalence calculated from known disease age/sex distributions and the practice age/sex distribution. In practices with a genuinely lower than expected disease prevalence, the stability of the register despite the regular application of Housekeeping procedures over time ( e.g. annually) will indicate completeness.

In the light of the above considerations a definition of a complete register is proposed with a grading system for incomplete ones. 

A complete register is one where (Number on disease register)/ (Expected number on disease register)>0.9 or where this ratio is <0.9 the register is stable (i.e. changing at <+/-5% of its prevalence p.a. where the disease prevalence is known to be stable). 

The choice of 0.9 is arbitrary but it seems reasonable to expect disease registers to be at least 90% of their expected value. It also allows for some variation in true disease prevalence between practices. The experience of housekeeping so far suggests that it rapidly brings registers nearer to the 90% mark for diabetes. It may be possible to make this definition more sophisticated as practices with complete registers become more common. We will then have a better idea of the true variation in disease prevalence between practices and will be able to adjust the expected value for factors like ethnic or social deprivation, rather than just basing it on the age/sex distribution. 

Practice disease registers can be graded P or E 1,2,3 or 4 as follows: 

APPENDIX 3  

Electronic Register 

Housekeeping Miquest Searches 

• C10 code present but no confirmatory sub code for diabetes i.e. no C10E/ C10F. 

• C10 code absent but sub code for diabetes present eg code for diabetic neuropathy. 

• C10 code absent but on diabetes drugs in the last year. 

• Patients with codes for both Type I and Type II diabetes in their records.

• For earliest C10E code (Type I) present other earlier codes for diabetes are present i.e. earliest C10E is not date of diagnosis. 

• For earliest C10F code (Type II) present other earlier codes for diabetes are present i.e. not date of diagnosis. 

• C10 code present but no code for locus of care i.e. 66AQ and 66AP absent for General Practice or hospital follow-up.