ARIA - Ratings explained

1. Introduction
Two panels consisting of 5 General Practitioners, 5 Cardiologists and 2 cardio-thoracic surgeons met in Summer 2003 to rate 1128 indications (combinations of clinical and patient characteristics) for appropriateness of Exercise testing and 3071 indications for coronary angiography.

Each indication was rated on a scale from 1 to 9, 1-3 denoting inappropriate, 4-6 equivocal and 7-9 appropriate.

The judgements of the panellists were based on the balance of clinical benefit : harm for each individual patient constellation based on the best available research evidence and their own expertise. It was assumed that cost criteria did not apply, there were no waiting lists and every clinician had immediate open access for each investigation. These assumptions were made because the ratings should capture the clinical necessity of an investigation that is applicable universally and over time. Factors such as access restrictions or waiting lists are considerations that a clinician has to take in the context of clinical appropriateness. Sometimes this will mean that investigations cannot be carried out even though the patient remains clinically appropriate.

For the same reason the patient was assumed to be free of major comorbidity, absolute indications, absolute contra-indications or preference. The presence of any of these factors may justifiably override any appropriateness for exercise ECG and coronary angiography. A patient who refuses an angiogram will not receive it despite the potential benefit of angiography.

2. Factors thought to influence appropriateness of testing
An indication is defined by several clinical descriptors such that each patient satisfying one indication is equally appropriate (or inappropriate) for a given management step.

1 Previous abnormal coronary angiogram.
This is defined as the presence of at least one vessel with at least a 50% stenosis.

2 Previous normal coronary angiogram.
This is defined as absence of any luminal narrowing.

3. Previous acute coronary syndrome (acute myocardial infarction or unstable angina)
Defined on the basis of typical history and electrocardiographic changes or biochemical changes. Thus the acute coronary syndrome is “definite,” not based on history alone. Thus in these patients the diagnosis of underlying coronary disease is in little doubt.

4. Typical angina symptoms

5. Atypical anginal symptoms
Pain or discomfort accompanied by any 2 out of 3

6. Non-specific chest pain
Pain or discomfort accompanied by any 1 out of 3
N.B. This widely used definition, does not include time to relief, or nature of chest pain.

Ten specific factors defining indications

Within each of the 6 broad clinical presentations, specific indications are defined using some or all of the following clinical descriptors.
1. Previous revascularisation. Either PCI or CABG, assumed to be complete revascularisation.
2. Time since angiography. Self explanatory
3. Time since acute coronary syndrome. Self explanatory.
4. Age. A maximum of 5 levels is considered (<40, 40-49, 50-59, 60-74, 75-84)
5. Sex
6. Severity of symptoms. Canadian Cardiovascular Society (CCS) scale is used for defining angina severity71. The CCS class is defined according to the box. In practice panellists may choose to consider a more general classification of mild (CCS class I-II), moderate (III) or severe (IV). The severity of symptoms is current, once the patient is on anti-anginal medication.

Functional severity of non-specific chest pain
For patients with non-specific chest pain it makes no sense to describe a CCS class. However such patients do differ in their functional limitations from their pain. For the purposes of rating this is defined at two levels as mild – moderate or severe. This mirrors clinical judgement in which some patients present with non-specific chest pain have severe functional impairment, or represent an important clinical problem because of repeated admissions.

7. Risk factors
Those factors which add incremental diagnostic or prognostic information to knowledge of age, sex and clinical presentation are termed “risk factors”.

As the evidence commentary shows, there are no scores validated in primary care which demonstrate how simple risk factors combine to contribute incremental diagnostic or prognostic value. The Framingham risk scores are not relevant because people with angina at baseline were excluded from the cohort.

Risk is defined simply, and partly implicitly, as low, medium or high based on smoking, hypertension, lipids, family history, obesity, south Asian ethnicity, diabetes and arterial disease in non-coronary circulations. Note age and sex are considered separately. Clinicians vary in which combination of risk factors they denote as “high risk” but for purposes of standardising the panel discussion the following definitions should be used:

Low risk “0” risk factors
Medium risk 1-2 risk factors
High risk 3 or more risk factors or
diabetes or
proven arterial disease in cerebral or peripheral circulations

8. Resting ECG abnormality
Is defined by the presence of pathological Q waves, or regional ST depression, or regional T wave inversion. These resting ECG abnormalities are considered to be ischaemic in origin.

9. Exercise ECG
All exercise ECG results are current i.e. the test was carried out at the time of the current presentation with symptoms. The following definitions have been used by RAND,72 and the ACRE study13 in previous expert panels rating the appropriateness of coronary angiography.

a. None
when no current exercise ECG results are available

b. Normal
The absence of a very abnormal or abnormal test (see below) and any one of the following achieved:
· at least 85 percent of the predicted max­imum heart rate
· heart rate/blood pressure product (heart rate x systolic arterial pressure + 100) of 250 greater
· 10 METS
· completed stage IV

c. Abnormal
After the first 3 minutes of the test the patient develops
(i) 1 mm or more of horizontal or downsloping ST segment depression that is present 80 msec after the J-point or
(ii) typical angina

d. Very abnormal
(i) During the first 3 minutes of the test (or onset at heart rate less than 120 beats/minute, off beta-blockers, or less than 6.5 METS) the patient develops
§ 1 mm or more of hori­zontal or downsloping ST segment depression that is present 80 msec after the J-point or
§ typical angina or
(ii) Based on blood pressure criteria, either of the following
§ exercise test stopped because of a fall in blood pressure or
§ test read as abnormal and at least two consecutive 10mm blood pressure systolic drops from the baseline (defined as the lower of standing or hyperventilation systolic blood pressure) which occurs during the exercise phase of the test.or
(iii) More than 2 mm of horizontal or downslop­ing ST depression at any time or
(iv) Persistence of ST depression greater than 6 minutes post exercise

N.B. An indeterminate test is not separately considered in the indications, on the grounds that it is equivalent to not done.

10. Medication for symptoms is defined as either sub-maximal or maximal. Sub-maximal is defined as only one agent, or two agents at sub-maximal dose. Maximal is defined as two or more agents at maximal dose, such that the physician has no further therapeutic option with drugs alone in the control of symptoms.

3. Ratings process
The RAND – Delphi method of rating the appropriateness of indications for referral and investigation of involve two rounds the first, in private, and the second where the panel members meet. All 22 panellists received software, ratings protocol and a review of the best available evidence and rated indications in private.

The software for rating has been developed by the ARIA investigators. The layout of the software is shown in Figure 1.

Figure 1 ARIA Wizard. Examples of indications and ratings: no previous ECG

After returning the ratings the scores from each panel member were collated for each indication and a personalised report on paper was generated. This showed the panellists own rating, the median of the whole panel, the range of the whole panel or whether the panel on the whole agreed or disagreed on an indication. Disagreement for an indication-decision pair is defined when:

Median of the panel = appropriate, but at least 2 panellists rate it inappropriate
or
Median of the panel = inappropriate, but at least 2 panellists rate it appropriate.

Panellists also received an agenda which highlighted the main areas for discussion at the two day panel meeting. The discussion focussed on important areas of disagreement and important areas of agreement, if the panels were grossly at odds with existing guidelines. The discussion aimed to achieve consensus and were asked to rerate indications if they were persuaded by the discussion. This process also helped eliminate any discrepancies due to misinterpretation or misunderstanding.

These ratings present the first appropriateness ratings for investigation in patients with angina by two independent panels.

This page last modified 21 October, 2005 by Web Administrator