DIABETES SCREENING DATABASE

A Comprehensive Electronic Patient Record for Global Risk Assessment in a

Rural Community

E. Pecoul

and H. F. Jelinek

IUP Génie Physiologique et Informatique, Faculté des Sciences Fondamentales Appliquées

Université de POITIERS, France

School of Community Health, Charles Sturt University, Albury, Australia

Keywords: Diabetes Screening, Access database, interdisciplinary health care.

Abstract: Interprofessional health care is becoming more prevalent with an increase in chronic diseases such as

diabetes and cardiovascular disease. In addition preventative models often require large numbers of risk

factors for identification of preclinical cases. CSU has established a diabetes screening clinic augmented by

an ACCESS database. The novelty of our work is that the ACCESS database integrates into the public

health sector and provides a more comprehensive review of health/disease indicators. Information on

traditional health indicators in addition to autonomic nervous system function tests, fundus examination and

foot assessment results as well as pro-inflammatory, pro-coagulation and antioxidant biochemistry can be

added by the university-based screening clinic as well as by diverse primary health care practitioners and

specialists that would otherwise not have access to this detailed information for patient assessment and

treatment. Our results are in favour of this interdisciplinary database indicating that over one year we have

identified 16.2% of people with no previous medical condition to have pre-diabetes, 2% had retinal disease

and 21.5% had foot problems. Moderate to severe ECG anomalies were identified in 19.3% of the

participants. Of these, 68.8% were either commenced on treatment, had their treatment changed or received

surgery. Our results indicate that a comprehensive EPR manager as part of an interdisciplinary health

screening initiative is able to track people that require intervention but were missed in the current public

health system as implemented in our community.

1 INTRODUCTION

For optimal health care, providers need the

necessary

information when they give treatment.

Health care provider and patient information and

decision

support needs can be satisfied if primary

care providers

use electronic patient records that are

comprehensive and provide the necessary

information to the general practitioner or specialist

at the time of consultation (EPRs) (Bates et al.,

2003). Often electronic records may indicate that a

patient has attended a specialist for an assessment

but the results may not be available to the general

practitioner nor to other primary health care

providers and specialists. In addition many rural

areas experience a lack in specialists and allied

health professionals as well as general practitioners.

Electronic patient records (EPRs) provide a

powerful opportunity for health assessments yob re

integrated as part of an integrated interdisciplinary

primary health network. Universities with additional

resources in rural communities can also to support

this initiative by establishing screening clinics as

part of their health professional courses. University-

based teaching and research can provide

screening/assessment results especially for tests not

routinely carried out in the community due to lack of

available services. With additional risk assessment

results during consultation, the GP's EPR is valuable

in identifying patients at risk for undiagnosed type 2

diabetes, cardiovascular disease or other illness.

Previous research on use of electronic patient

records (EPR) indicated that general practitioners

are able to utilise information on additional risk

factor assessment from the EPR during regular

consultation (Klein Woolthuis et al., 2007). EPR

recall systems based on HbA1c levels showed that

the median practice-specific proportions of patients

Pecoul E. and F. Jelinek H. (2008).

DIABETES SCREENING DATABASE - A Comprehensive Electronic Patient Record for Global Risk Assessment in a Rural Community.

In Proceedings of the First International Conference on Health Informatics, pages 13-19

 SciTePress

with HbA1c recorded annually increased from a

median 65% to 77% when using an EPR (Mohiddin

et al., 2006). An Australian study conducted by

Burns et al. using the CARDIAB®

system evaluated the process of care of patients for

nine parameters critical to quality of

care in diabetes:

glycaemic control (HbA1c), blood pressure,

body

mass index, total cholesterol levels,

triglyceride

levels, high density lipoprotein

levels (HDL),

microalbumin level, foot checks and eye checks.

They demonstrated the potential

of CARDIAB®

to provide support to

general practice in the

management of diabetes (Burns et al., 2004).

Divisions of General Practice, universities and

other primary health care providers have

a key role

to play in supporting general

practitioners (GPs) to

provide proactive, preventive care for

their patients

with cardiovascular disease

(CVD) and diabetes

(Penn et al., 2004). They can achieve this by

providing them with global risk marker information

and health assessment results. Risk stratification is

currently based on the Framingham cardiovascular

risk equation and therefore include factors such as

blood pressure, diabetes, age, gender and HDL

(Sheridan et al., 2003). However research has

suggested that a more global approach to

identification of diabetes and cardiovascular disease

may be appropriate (Michos et al., 2006). In addition

the risk status of persons

without CHD varies

greatly, and this variability mandates a

range in the

intensity of interventions. Effective primary

prevention

thus requires an assessment of risk to

categorize patients for

selection of appropriate

interventions and a more comprehensive screening

/health assessment that includes additional factors

provided by combining the information obtained

from an interdisciplinary health care system and

available via an EPR (Jelinek et al., 2006,

Woodward et al., 2006).

The major and independent risk factors for CHD

are cigarette

smoking, elevated blood pressure,

elevated serum

total cholesterol and low-density

lipoprotein cholesterol (LDL-C),

low serum high-

density lipoprotein cholesterol (HDL-C), diabetes

mellitus, and advancing age. Other factors are

associated with increased risk for

CHD. These

include obesity, physical inactivity, elevated

homocysteine, pro-coagulation and pro-

inflammatory factors. An assessment of global risk

based on the summation

of all major risk factors can

be clinically useful for in identification of high-risk

patients who require immediate

attention and

intervention and changes in intervention based on

the global risk estimate (Grundy et al., 1999).

Health assessment and determination of risk

factors included in a global risk equation is often not

possible in rural communities where there is a lack

of general practitioners, allied health practitioners

and specialists. University-based teaching and

research has the opportunity to integrate with the

public health system (Jelinek et al., 2006). An

electronic patient record system that spans from

university-based teaching and research to public

health care and that incorporates data obtained from

an interprofessional health care system is required.

The EPR must integrate medical, nursing, and

allied health worker data as well as the results from

pathology laboratories that may be shared by the

health care team. The EPM needs to be secure,

provide relevant template forms, incorporate a recall

system and specific reports, and a means of tracking

interventions by diverse health care professionals.

The majority of databases meet this criterion but are

mainly specific for certain medical specialities such

as general practice. The DiScRi database discussed

here extends these fundamental principles to address

the needs of an interdisciplinary health care team

with emphasis on identification and follow-up of

people with and without diabetes or cardiovascular

disease as well as identification of those at risk of

diabetes and cardiovascular disease based on a

global risk assessment. Information of outcomes

following referral to general practitioners, allied

health and medical specialists provides information

to of global health care and the status of the patient.

The database allows access to all biochemistry and

pathology results as well as the retinal photography,

12-lead ECG and Doppler peripheral vessel blood

flow velocity traces that is collected by diverse

health care professionals that would otherwise not be

available in a coherent form.

2 METHODS

2.1 Database

The Diabetes Screening Research Initiative (DiScRi)

application was designed using Microsoft ACCESS

(under Microsoft Windows), SQL (Select Query

Language) and VBA (Microsoft Visual Basic)

languages. The CDM (Conceptual Data Model) and

LDM (Logical Data Model) were designed using

MERISE theories. The database contained 22

entities and 153 attributes (Table 1).

Accurate information is available for

demographics, medical history, biochemistry, urine

analysis as well as lifestyle, medication, foot, eye

HEALTHINF 2008 - International Conference on Health Informatics

and heart health including antioxidant levels, pro-

inflammatory and pro-coagulation factors for a

patient. There is also a direct access to files such as

retinal and ECG pictures, and a resume of most

important health information (BP, age, cholesterol,

diabetes status, medication).

The body-mass index (BMI), the ankle-brachial

pressure index (ABPI) for assessment of peripheral

vascular disease, the Ewing score, which indicates

presence and severity of cardiac autonomic

neuropathy as well as absolute 5 year CVD risk are

automatically calculated. Cover letters for the

general practitioner or specialist, if required, are

automatically generated with a link between

Microsoft Access and Microsoft Word (an update in

the database will update the Word document) and

can be sent by e-mail (using Word mail function), or

printed for mailing and/or storage as hard copy.

2.2 Screening Participants

To assess the utility of the DiScRi database, patients

were recruited through the media and the research

had Charles Sturt University Human Ethics

approval. All participants had their height, weight

and waist circumference measured. A urine sample

was analysed and a blood sample used to determine

levels of glucose, electrolyte, cholesterol, HbA1c,

antioxidants, pro inflammatory and pro-coagulation

markers. Retinal photographs of the posterior pole

are used to determine presence of diabetic

retinopathy and risk of CVD. Heart function was

determined by 12-lead ECG including calculation of

corrected QT interval (QTc), QT dispersion (QTd),

frequency and time domain analysis results. The

latter are indicators of preclinical cardiac autonomic

neuropathy (CAN). CAN is also assessed using the

Ewing battery. Foot health assessment includes

ABPI, monofilament for peripheral neuropathy

(PN), ankle and knee reflexes, and muscle tone.

Thus the university screening simulated an

interdisciplinary health provision system.

3 RESULTS

3.1 Database Characteristics

From the introductory screen, the database allows

access the demographic data, clinical history,

follow-up, biochemistry, eye, heart and foot

assessment results.

Figure 1 shows an example of a test recording

screen. For the ECG analysis, the 12-lead ECG trace

can also be accessed. The 3-lead recording indicates

the results for the assessment of cardiac autonomic

neuropathy based on the Ewing score.

All data is divided into tables that can be

Table 1: Entities contained in the DiScRi database.

Entities # values Examples

GP 4 Name, address, telephone, email

Patient 5 Name, address, telephone, DOB

Analysis 2

Medications 2 Current medications, change medications

Clinical history 5 Diabetes, HT, CVD, medical information

Foot history 3 Presence of ulcers, general pain, numbness

Heart history 10 Angina, fainting, heart attack, stroke, chest pain

Eye history 2 Glaucoma, cataract, diabetic retinopathy

Other history 6 Kidney, bladder, GIT

Lifestyle 7 Smoking, alcohol, family history for diabetes/CVD, diet, exercise

Consultation history 6 Attendance to health care professionals; details of consultations

3lead ECG Ewing 7 Lying to standing HRV change

3lead ECG HRV 17 Frequency and time domain results for HRV

12lead ECG 8 HR , QTc, QTd, interpretation,

Neuropathy 7 Monofilament, ABPI,

BMI 4 Height, weight, waist circumference

ABPI 5 Systolic blood pressure at arm and ankle

Glucose 5 time since last meal, glucose level at screening

BP 5 Lying and standing BP

Urine 12 Glucose, protein, albumin

Biochemistry 26 Cholesterol, antioxidants, pro-inflammatory, pro-coagulation markers

Eye analysis 5 Eye pathology, AVR diagnosis, photograph

DIABETES SCREENING DATABASE - A Comprehensive Electronic Patient Record for Global Risk Assessment in a

Rural Community

separately accessed, However for a quick overview

during a consultation, the global information page

indicates the date of last visit and includes the

possibility of accessing the eye photographs of the

retinal posterior pole, the 12-lead ECG trace and

information on glucose and cholesterol levels as well

as blood pressure and BMI (Figure 2). From this

page the user can also move to any of the analysis

for specific and more comprehensive results.

An important addition to this database is the

opportunity for detail follow-up provided by diverse

health care practitioners as shown in Figure 3.

3.2 Screening Outcomes

A review of the outcomes for the university-based

simulation study that measured health/disease

indicators indicated that in this rural area the

prevalence is worse compared to the national

average for all tests in the diabetes cohort and

elevated for the ABPI and ECG in the non diabetes

group (Table 2).

Table 2: Prevalence of pathology in a rural cohort

determined through a university-based screening initiative.

Measure* % Non

diabetes

% National

Average

PN 8.13 19.4 (foot

ulcer)

ABPI 11.6 unknown

BMI 25 20

(>30kg/m2)

HT 28 30

(140/90)

ECG 13 12(diab) /

3(control)

Figure 1: ECG assessment form.

HEALTHINF 2008 - International Conference on Health Informatics

Figure 2: Global information screen with eye photograph.

Figure 3: Recall screen.

DIABETES SCREENING DATABASE - A Comprehensive Electronic Patient Record for Global Risk Assessment in a

Rural Community

4 DISCUSSION

In Australia, general practitioners play an important

role in the management of diabetes. This involves

careful monitoring of behavioural risk factors, blood

pressure, glycaemic control and lipids; early

detection of complications; and management

according to evidence-based guidelines (NHMRC,

2004, Newnham et al., 2004). However, a

comprehensive health review can only be provided

by an interdisciplinary health care team. With the

rapid advances in information technology in the last

decade, various diabetes information systems have

evolved in different parts of the world. Availability

of new technologies and information systems for

monitoring and treating diabetes is critical to

achieving recommended metabolic control. A

comprehensive EPR that includes data for global

risk assessment and patient review and provides

mechanisms for practitioners to gain information on

performance and results from a diverse primary

health care team that would otherwise not be

integrated is essential for evidence-based practice

and improvement in health care provision.(Joshy and

Simmons, 2006) The DiScRi study indicates that an

EPR that incorporates information across the health

care sector arising from annual consultations in the

university

setting such as test results and

outcome of

referrals, as well as lifestyle data such as smoking

status, exercise

and body mass index is of benefit to

both the health care providers and the patients.

Effective EPR implementation and

networking

could eventually save more than $81 billion

annually—by

improving health care efficiency and

safety—and that HIT-enabled

prevention and

management of chronic disease could eventually

double those savings while increasing health and

other social

benefits.(Hillestad et al., 2005) DiScRi

data provides important baseline information

for

health care quality improvement at local, state and

national levels. Including pro-coagulation and pro-

inflammatory factors improves overall health care

planning as the physician has additional

information.(Navab et al., 2006) Similarly, options

for viewing retinal images that indicate early signs

of diabetes or presence of CVD seen on 12-lead

ECG or peripheral pulse wave analysis can be of use

to general practitioners. The database allows access

to all biochemistry and pathology results as well as

the retinal photography, 12-lead ECG and Doppler

peripheral vessel blood flow velocity traces. Making

the DiScRi EPR the most comprehensive primary

health care database available in Australia.

REFERENCES

Bates, D. W., Ebell, m., Gotlieb, E., Zapp, J. & Mullins,

H. C. (2003) A Proposal for Electronic Medical

Records in U.S. Primary Care. J Am Med Inform

Assoc, 10, 1-10.

Burns, J. R., Bonney, M.-A., Davies, P. G. P. & Harris, M.

F. (2004) Utilization of the CARDIAB(R)TM

database system to promote quality of care in

Australian general practice. Health Informatics

Journal, 10, 177-184.

Grundy, S. M., Pasternak, R., Greenland, P., Smith, S., Jr.

& Fuster, V. (1999) Assessment of Cardiovascular

Risk by Use of Multiple-Risk-Factor Assessment

Equations : A Statement for Healthcare Professionals

From the American Heart Association and the

American College of Cardiology. Circulation, 100,

1481-1492.

Hillestad, R., Bigelow, J., Bower, A., Girosi, F., Meili, R.,

Scoville, R. & Taylor, R. (2005) Can Electronic

Medical Record Systems Transform Health Care?

Potential Health Benefits, Savings, And Costs. Health

Aff, 24, 1103-1117.

Jelinek, H. F., Wilding, C. & Tinley, P. (2006) An

innovative multi-disciplinary diabetes complications

screening programme in a rural community: A

description and preliminary results of the screening.

Australian Journal of Primary Health, 12, 14-20.

JOSHY, G. & SIMMONS, D. (2006) Diabetes

Information Systems: A Rapidly Emerging Support

for Diabetes Surveillance and Care. Diabetes

Technology & Therapeutics, 8, 587-597.

Klein Woolthuis, E. P., De Grauw, W. J. C., Van Gerwen,

W. H. E. M., Van Den Hoogen, H. J. M., Van de

Lisdonk, E. H., Metsemakers, J. F. M. & Van Weel, C.

(2007) Identifying people at risk for undiagnosed type

2 diabetes using the GP's electronic medical record.

Fam. Pract., 24, 230-236.

Michos, E. D., Nasir, K., Braunstein, J B., Rumberger, J

A., Budoff, M. J., Post, W. S. & Blumenthal, R. S.

(2006) Framingham risk equation underestimates

subclinical atherosclerosis risk in asymptomatic

women. Atherosclerosis, 184, 201-206.

Mohiddin, A., Naithani, S., Robotham, D., Ajakaiye, O.,

Costa, D., Carey, S., Jones, R. H. & Gulliford, M. C.

(2006) Sharing specialist skills for diabetes in an inner

city: A comparison of two primary care organisations

over 4 years. Journal of Evaluation in Clinical

Practice, 12, 583-590.

Navab, M., Anantharamaiah, G. M., Reddy, S. T., Van

Lenten, B. J., Ansell, B. J. & Fogelman, A. M. (2006)

Mechanisms of disease: proatherogenic HDL - an

evolving field. National Clinical Practice,

Endocrinology and Metabolism, 2,

504-511.

Newnham, H., Colagiuri, S., Hepburn, A., Chen X. M. &

Colagiuri, R. (2004) National Evidence Based

Guidelines for Type 2 Diabetes: Prevention and

Detection of Macrovascular Disease. Canberra,

NHMRC.

HEALTHINF 2008 - International Conference on Health Informatics

NHMRC (2004) National Health and Medical Research

Council. National evidence based guidelines for the

management of type 2 diabetes mellitus: prevention

and detection of macrovascular disease. Sydney,

Australian Centre for Diabetes Strategies.

Penn, D. L., Burns, J. R., Georgiou, A., Davies, P. G. P. &

Harris, M. F. (2004) Evolution of a register recall

system to enable the delivery of better quality of care

in general practice. Health Informatics Journal, 10,

165-176.

Sheridan, S., Pignone, M. & Mulrow, C. (2003)

Framingham-based Tools to Calculate the Global Risk

of Coronary Heart Disease. A Systematic Review of

Tools for Clinicians. Journal of General Internal

Medicine, 18, 1039-1052.

Woodward, A., Wallymahmed, M., Wilding, J. & Gill, G.

(2006) Successful cardiovascular risk reduction in

Type 2 diabetes by nurse-led care using an open

clinical algorithm. Diabetic Medicine, 23, 780-787.

DIABETES SCREENING DATABASE - A Comprehensive Electronic Patient Record for Global Risk Assessment in a

Rural Community