INTERCHEST Clinical Prediction Rule for Chest Pain in Primary Care (Calculator)

INTERCHEST Prediction Rule Explained

This is a clinical predictor that rules out coronary artery disease (CAD) in primary care patients 30 years and older, presenting with chest pain. This is not a tool for the emergency setting. If an obvious cause of chest pain can be established, such as trauma or infection, or anginal symptoms, this prediction rule is not recommended.

The purpose of the predictor is to screen for patients with low risks and CAD unlikely, so that they can be discharged without further evaluation.

The INTERCHEST cardiac pain rule has shown better predictive properties than the Marburg Hearst score but still requires further validation studies.

Diagnostic risk stratification scores for chest pain may also improve telephone triage for major events in out-of-hours primary care, by reducing the number of unnecessary referrals without compromising triage safety.

Scoring and Interpreting the INTERCHEST

INTERCHEST scores of ≤1 in primary care indicate a very low likelihood of chest pain due to unstable CAD (98% NPV), making further non-urgent outpatient evaluation safe.

Patients with an INTERCHEST score of ≤1 typically don't need urgent evaluation unless they show clear clinical instability; they can be managed through routine outpatient follow-up.

Patients with an INTERCHEST score of ≥2 don't necessarily have unstable CAD but should undergo more urgent testing.

The final score is computed by adding the points together. These are then interpreted in reference to a cut-off on 1 point. Scores of 0 or 1, indicate that unstable CAD is highly unlikely (NPV 98%) whilst scores of 2 or more are modestly predictive of CAD (PPV 43%).

A patient with a high risk INTERCHEST score (≥2) might not have unstable CAD, but since this cannot be ruled out if the score is ≥3, these patients usually need more urgent evaluation or inpatient admission.

If there are obvious symptoms resembling angina (like jaw pain, shortness of breath during activity, or arm pain) or if the ECG shows ischemic changes, the INTERCHEST score is irrelevant and urgent inpatient admission is necessary.

About the Original Study and Further Studies

The objective of the study by the international group and collaborators, was to create a clinical prediction rule for diagnosing coronary artery disease (CAD) in primary care patients presenting with chest pain.

Data from 3,099 patients across five studies were used, incorporating methods like random forest trees, multiple imputation of missing values, and logistic regression to identify six key weighted predictors.

The prediction rule, which was generated using these factors, demonstrated an area under the ROC curve of 0.84, indicating good diagnostic ability. Applying this rule to a dataset with a 13.2% prevalence of CAD, using a prediction score cutoff of <2 (i.e., -1, 0, or +1), resulted in a CAD probability of 2.1% (95% CI: 1.1-3.9%) when the score was <2, and 43.0% (95% CI: 35.8-50.4%) when the score was ≥2. CAD was considered absent with a prediction score <2.

The study emphasizes the importance of using large data sets from diverse sites to improve the internal and external validity of clinical prediction rules.

A further study reviewed clinical decision rules (CDRs) used in primary care for assessing intermittent chest pain and ruling out acute coronary syndrome (ACS) without advanced diagnostic testing. Eight studies were analyzed, covering five CDRs: Gencer rule, Marburg Heart Score, INTERCHEST, Grijseels rule, and Bruins Slot rule.

Three CDRs (Gencer, Marburg Heart Score, INTERCHEST) were designed to rule out coronary disease, with the Marburg Heart Score showing the highest methodological quality. It demonstrated consistent sensitivity (86%-91%), specificity (61%-81%), and predictive values (PPV: 23%-35%, NPV: 97%-98%). The Gencer rule and INTERCHEST had comparable diagnostic performance but require further validation.

Two CDRs (Grijseels, Bruins Slot) focused on ruling out ACS. The Grijseels rule had a sensitivity of 91% and specificity of 37%, while the Bruins Slot rule had a sensitivity of 97% and specificity of 10%. The Marburg Heart Score outperformed clinical judgment in detecting coronary disease, while the Bruins Slot rule was deemed safer than clinical judgment alone, despite a limited sample size.

The study concluded that no clinical decision aid can safely rule out ACS in general practice, though several rules exist for intermittent chest pain, with the Marburg Heart Score being the most reliable.

References

Original reference

International Working Group on Chest Pain in Primary Care (INTERCHEST); Aerts M, Minalu G, Bösner S, Buntinx F, Burnand B, Haasenritter J, Herzig L, Knottnerus JA, Nilsson S, Renier W, Sox C, Sox H, Donner-Banzhoff N. Pooled individual patient data from five countries were used to derive a clinical prediction rule for coronary artery disease in primary care. J Clin Epidemiol. 2017; 81:120-128.

Other references

Manten A, De Clercq L, Rietveld RP, Lucassen WAM, Moll van Charante EP, Harskamp RE. Evaluation of the Marburg Heart Score and INTERCHEST score compared to current telephone triage for chest pain in out-of-hours primary care. Neth Heart J. 2023; 31(4):157-165.

Sox HC, Aerts M, Haasenriter J. Applying a Clinical Decision Rule for CAD in Primary Care to Select a Diagnostic Test and Interpret the Results. Am Fam Physician. 2019; 99(9):584-586.

Cayley WE Jr. Diagnosing the cause of chest pain. Am Fam Physician. 2005; 72(10):2012-21.

Klinkman MS, Stevens D, Gorenflo DW. Episodes of care for chest pain: a preliminary report from MIRNET. Michigan Research Network. J Fam Pract. 1994; 38(4):345-52.