|
Clinical Prediction
Rule Estimates
Clinical prediction
rules describe how to use the key clinical findings that predict
a disease to estimate the pre-test probability of disease in a
patient.
For example, a chest
pain rule may use regression analysis of a standardized set of
data, including clinical findings (the predictor variables) and
the final diagnosis (the dependent variable) to identify the best
clinical predictors and their diagnostic weights. The sum of the
diagnostic weights corresponding to a patient's findings is a
score, and the probability of disease for each patient is equivalent
to the prevalence of disease among patients with a similar score.
The table below displays
a chest pain rule to estimate the pre-test probability of a positive
coronary arteriogram.
| Table
1.5.2: Chest Pain Rule to Estimate Pre-test Probability
|
|---|
| Attribute
| Coefficient
| RoundedCoefficient
|
|---|
| Age
>60 years | +2.85
| +3 |
Pain
is exertional
| +4.26
| +4 |
| Pain
causes patient to stop all activities | +2.76
| +3
|
| History
of MI | +3.90
| +4
|
| Pain
relieved w/i 3 minutes by NTG | +1.93
| +2
|
| >
20 pk-yrs smoking | +3.93 |
+4
|
| Male
gender | +5.37 |
+5 |
| Reference:
Sox HC, Hickam DH, Marton KI, et al. Using the patient's
history to estimate the probability of coronary artery
disease: a comparison of referral and primary care practice.
Am J Medicine. 1990;89:7-14. |
|
Question 1.5.3 - Applying the Method
What is the chest pain
score of the case study patient?
Click
to review patient history
 | Chest
pain score = 5 |
 | Chest
pain score = 7 |
 | Chest
pain score = 8 |
 | Chest
pain score = 12 |
 | Chest
pain score = 15 |
 | Chest
pain score = 19 |
Research Opportunities:
Clinical Prediction Rules
Despite their usefulness,
relatively few symptoms have clinical prediction rules, so the
door is open, one might even say wide open, for investigators
interested in establishing these tools.
Keep in mind that clinical
prediction rules must:
- Describe
the key clinical findings that predict a disease, and
- Show
how to use these findings to estimate the probability of disease
in a patient.
|