Newsom Score for non-traumatic chest pain

Clinical background

Chest pain is one of the commonest symptoms in the emergency department, and a large proportion of these patients undergo a screening chest radiograph as a matter of routine. The imaging is, however, of low yield in most cases: the clinical yield of radiographic findings requiring action lies between 2 and 6 per cent in this population [1, 3]. The Newsom score for non-traumatic chest pain was developed to identify the subgroup of patients in whom a radiograph can reasonably safely be omitted, and thereby to reduce unnecessary radiation, waiting times and use of resources.

The tool is not a diagnostic instrument for acute coronary syndrome, pulmonary embolism or aortic dissection. It is designed to detect five specific radiographic findings: pneumonia, pleural effusion, pneumothorax, heart failure and a new tumour [1]. The decision the tool serves is therefore narrow: should a screening chest radiograph be ordered or not, given that the patient has already been assessed for acute coronary syndrome and that this assessment points to a low probability?

Applying the Newsom score

The Newsom score is a binary checklist of twelve variables. Each variable scores 0 if absent and 1 if present. The score is not additive in the traditional sense but functions as a threshold tool:

$$\text{Newsom} = \mathbf{1}\left[\sum_{i=1}^{12} x_i \geq 1\right]$$

where $x_i$ is the presence (1) or absence (0) of the $i$th variable. If at least one variable is positive, the patient is classified as not low risk and a chest radiograph is recommended. If all twelve variables are negative, a low-risk group is identified in which a clinically significant radiographic finding is unlikely.

The twelve variables are:

The derivation cohort consisted of 1,111 adult patients presenting to three emergency departments in the USA with non-traumatic chest pain as the main symptom [1]. Of these, 1,089 chest radiographs were analysed. The study was prospective and observational: the emergency physician completed a standardised data form with history and examination variables before the radiograph was ordered. The outcome measure was the presence of a clinically significant radiographic finding, defined as pneumonia, pleural effusion, pneumothorax, heart failure or a new tumour. Disease prevalence was low: 70 patients (6.4 per cent) had a clinically relevant finding [1].

The rule is a development of earlier work. In 2010 Hess et al. derived a simpler rule with three variables (history of heart failure, smoking history, auscultatory findings) that showed 100 per cent sensitivity in the derivation cohort [2]. That rule failed on prospective validation by Poku et al. in 2012, in which sensitivity fell to 78.3 per cent [3]. The Newsom rule expands the set of variables to twelve factors in order to improve sensitivity at validation.

Interpretation in practice

The score is thus not intended to grade risk in levels. It is a binary "omit or not" rule: the patient either meets the criteria for forgoing the radiograph or does not. There is no middle ground in which a higher score would warrant different management from a lower one.

A patient falling into the low-risk group should still be assessed clinically for acute coronary syndrome according to usual practice (ECG, biomarkers where needed). The Newsom score does not replace that work-up; it complements it by answering one specific question: is a chest radiograph needed?

Validation and performance

The only prospective validation of the Newsom rule is the derivation study itself, that is, the same cohort in which the rule was refined [1]. No independent external validation has been published. The authors themselves state explicitly that further validation is needed before the rule can be recommended for clinical use [1].

In the validation cohort the rule performed as follows [1]:

• Sensitivity: 92.9 per cent (95 per cent CI 83.4 to 97.3)
• Specificity: 30.4 per cent (95 per cent CI 27.6 to 33.4)
• Negative predictive value: 98.4 per cent (95 per cent CI 96.1 to 99.4)

The rule missed five patients with clinically significant radiographic findings: three pneumonias and two pleural effusions [1]. This corresponds to a false negative rate of 7.1 per cent of all patients with genuine findings. Applying the rule would have reduced chest radiograph use by 28.9 per cent in the cohort [1].

The low specificity is expected and accepted: the rule is designed for high sensitivity at the expense of specificity, since missing a radiographic finding is judged more serious than performing a superfluous radiograph. A specificity of 30.4 per cent means that more than two in three patients without a finding will still undergo radiography, which limits the actual reduction in imaging.

A systematic review and meta-analysis from 2026 of clinical decision rules for chest radiography in non-traumatic chest pain identified seven studies with a total of 6,654 patients [4]. The review analysed two rules: the Hess rule and the Rothrock rule. The Newsom rule is not mentioned in the review, reflecting its lack of independent validation. For the Hess rule, a pooled sensitivity of 98.3 per cent and a specificity of 47.6 per cent were reported, but with high heterogeneity between studies [4]. The authors conclude that none of the existing rules can be recommended for clinical use, and call for prospective derivation according to established methodological standards [4].

Limitations

Lack of independent validation. This is the single most important limitation. The Newsom rule has been tested in one cohort only, by the same research group that developed it, at three American emergency departments. Generalisability to other populations, health systems or prevalence levels is unknown. The authors themselves do not recommend clinical use before external validation is available [1].

False negative cases. The rule missed five of 70 patients with genuine findings, three of them pneumonias [1]. A patient with pneumonia who lacks all twelve risk markers is possible, particularly among younger patients without comorbidity. Persisting clinical suspicion must always outweigh a low-risk result.

Low specificity. With a specificity of 30.4 per cent, radiograph use is reduced by barely 29 per cent [1]. This is considerably less than was achieved with the Canadian ACS rule in an Australian validation study, in which a reduction of 47 per cent was reported, albeit at a higher disease prevalence (12 per cent) and with a different rule structure [5].

A narrow outcome definition. The rule detects only five specific findings. Other radiographic findings that may be clinically relevant, for example mediastinal changes, a pericardial effusion or fractures, are not part of the outcome measure. A patient with aortic dissection may have a normal Newsom score, but this is not the purpose of the rule: it is not intended for the investigation of acute coronary syndrome or other life-threatening conditions.

Composition of the population. The cohort consisted of patients with non-traumatic chest pain in whom the emergency physician had already judged a chest radiograph to be indicated. The rule is not validated for patients in whom radiography is not considered from the outset, nor for patients with trauma, dyspnoea as the main symptom, or known lung disease with an acute deterioration.

Previous failures of similar rules. The predecessor Hess rule, which shares three variables with the Newsom rule, fell from 100 per cent sensitivity in its derivation to 78.3 per cent on independent validation [2, 3]. This illustrates the risk of overfitting in prospective validation studies carried out by the same group, and reinforces the requirement for independent validation before the Newsom rule can be considered clinically safe.

References

1. Newsom C, Jeanmonod R, Woolley W, et al. Prospective Validation and Refinement of a Decision Rule to Obtain Chest X-ray in Patients With Nontraumatic Chest Pain in the Emergency Department. Acad Emerg Med. 2018;25(6):650-656. PMID: 29427301
2. Hess EP, Perry JJ, Ladouceur P, et al. Derivation of a clinical decision rule for chest radiography in emergency department patients with chest pain and possible acute coronary syndrome. CJEM. 2010;12(2):128-134. PMID: 20219160
3. Poku JK, Bellamkonda-Athmaram VR, Bellolio MF, et al. Failure of prospective validation and derivation of a refined clinical decision rule for chest radiography in emergency department patients with chest pain and possible acute coronary syndrome. Acad Emerg Med. 2012;19(9):E1004-E1010. PMID: 22978726
4. Al Fathil Y, Rotella JA. Clinical decision rules for obtaining chest radiography in adult patients presenting to the emergency department with non-traumatic chest pain: a systematic review and meta-analysis. Emerg Med J. 2026. PMID: 42091207
5. Goldschlager R, Roth H, Solomon J, et al. Validation of a clinical decision rule: chest X-ray in patients with chest pain and possible acute coronary syndrome. Emerg Radiol. 2014;21(4):367-372. PMID: 24595499