Altitude-adjusted PERC rule

Clinical background

PERC (the Pulmonary Embolism Rule-out Criteria) was developed to identify emergency department patients whose probability of pulmonary embolism is so low that further investigation can be withheld, without the need for a D-dimer or computed tomography pulmonary angiography (CTPA). The rule is an exclusion tool and should be applied only after clinical assessment has already established a low probability, defined as a gestalt pre-test probability below approximately 15% [2].

One of the PERC criteria is an oxygen saturation below 95%. At high altitude, baseline oxygen saturation is physiologically lower, and a threshold of 95% causes an unnecessarily large number of patients to be misclassified as PERC-positive, driving further investigation without improving safety. The altitude-adjusted PERC rule lowers the threshold for the hypoxia criterion to an oxygen saturation below 90%, with the aim of reducing the number of false-positive classifications at institutions located at high altitude [1].

Applying the altitude-adjusted PERC rule

The rule consists of eight binary criteria, each assessed as yes or no:

$$\text{PERC} = \text{Age} \geq 50 + \text{HR} \geq 100 + \text{SpO}_2 < 90$$

Each criterion scores 0 (no) or 1 (yes). PERC is negative only if all eight criteria are zero. The only difference from standard PERC is that the hypoxia criterion uses an oxygen saturation below 90% instead of below 95%.

The other seven criteria are unchanged: age 50 years or more, heart rate 100/min or more, unilateral leg swelling, haemoptysis, recent surgery or trauma within 4 weeks requiring hospital care, previous pulmonary embolism or DVT, and hormone (oestrogen) use.

The derivation study for the altitude-adjusted version was conducted at the University of Utah Emergency Department, situated at an altitude of 1,518 metres (4,980 feet) [1]. The authors prospectively recruited a convenience sample of 3,024 patients presenting with chest pain and/or dyspnoea. The prevalence of pulmonary embolism was 1.9%. The outcome was a diagnosis of acute pulmonary embolism during the emergency department visit. The study compared standard PERC (SpO₂ below 95%) with the altitude-adjusted version (SpO₂ below 90%) with respect to sensitivity and the potential reduction in imaging.

The original PERC rule was derived by Kline and co-workers in 2004 and validated prospectively in a multicentre study at 13 US emergency departments with 8,138 patients [2]. In that cohort, 20% of patients had a low gestalt probability combined with a PERC-negative result, and the composite false-negative rate (VTE or death within 45 days) was 1.0% (95% CI 0.6 to 1.6), with a sensitivity of 97.4% and a specificity of 21.9% [2].

Interpretation in practice

It is important to understand that a positive PERC does not mean a high probability of pulmonary embolism. It means only that the exclusion strategy is not applicable. The majority of PERC-positive patients do not have pulmonary embolism, but they require further investigation before the diagnosis can safely be dismissed.

The altitude-adjusted version is intended only for institutions at high altitude, where the normal population has a lower baseline oxygen saturation. At sea level, standard PERC with the 95% threshold should be used.

Validation and performance

In the derivation study for the altitude-adjusted version, the sensitivity of standard PERC was 96.6% (95% CI 88.1 to 99.6%) and that of the altitude-adjusted version 94.8% (95% CI 85.6 to 98.9%) [1]. The difference in sensitivity was small and the confidence intervals overlap. The altitude-adjusted version would have reduced the proportion undergoing advanced imaging by 2.7% (95% CI 1.8 to 4.1%) compared with standard PERC, assuming that PERC-negative patients did not undergo CTPA [1].

Extensive validation data exist for standard PERC. In the PROPER trial, a cluster-randomised crossover non-inferiority study at 14 French emergency departments with 1,916 patients at very low clinical probability, the rate of thromboembolic events at 3 months was 0.1% in the PERC group compared with 0% in the control group, which met the non-inferiority margin of 1.5% [3]. CTPA use fell from 23% in the control group to 13% in the PERC group (difference −10 percentage points, 95% CI −13 to −6) [3].

A meta-analysis from 2026 including 10 studies and 13,672 patients with a pooled pulmonary embolism prevalence of 7% reported a pooled sensitivity of 95% (95% CI 89 to 98%), specificity of 26% (95% CI 16 to 40%) and negative predictive value of 98.2% (95% CI 97.8 to 99.0%) for PERC [4]. Use of PERC was associated with a significant reduction in imaging (RR 0.85, 95% CI 0.80 to 0.91) [4].

An earlier meta-analysis by Singh and co-workers, based on 12 cohorts and 13,885 patients, found a pooled sensitivity of 97% (95% CI 96 to 98%) and specificity of 23% (95% CI 22 to 24%), with a negative likelihood ratio of 0.17 (95% CI 0.13 to 0.23) [5].

The altitude-adjusted version, however, has been validated in only a single study and lacks external validation. The figures for the altitude-adjusted variant derive exclusively from the Utah cohort of Madsen and co-workers [1].

Limitations

PERC applies only to patients with a low pre-test probability. It is an exclusion tool, not a tool for supporting a diagnosis of pulmonary embolism. Applying PERC to patients with a moderate or high clinical probability is incorrect and may lead to missed diagnoses. In a Swiss validation study at six emergency departments with a pulmonary embolism prevalence of 21.3%, the prevalence of pulmonary embolism among PERC-negative patients was 5.4% overall and 6.4% among those with a low pre-test probability, far exceeding the accepted safety margin [6]. The negative likelihood ratio in the low-risk group was 0.63 (95% CI 0.38 to 1.06), which is not good enough to exclude pulmonary embolism safely in a high-prevalence population [6].

The altitude-adjusted version has specific limitations:

• It was validated at an altitude of 1,518 metres. There are no data for other altitudes, and it is unclear whether the 90% threshold is appropriate at, for example, 2,500 metres or higher.
• The study used a convenience sample rather than consecutive enrolment, which may introduce selection bias.
• Only one study exists; there is no external validation.
• The study included patients with chest pain and/or dyspnoea, not all patients with suspected pulmonary embolism.

PERC does not apply to patients with another serious concurrent illness that explains the symptoms better than pulmonary embolism, nor to patients in whom clinical assessment cannot classify the probability as low. The rule should not be used in patients under 18 years, in pregnancy, or in patients with active cancer, as these populations were not well represented in the derivation cohorts.

References

1. Madsen T, Jedick R, Teeples T, Carlson M, Steenblik J. Impact of altitude-adjusted hypoxia on the Pulmonary Embolism Rule-out Criteria. Am J Emerg Med. 2019;37(2):281-285. PMID: 29848460
2. Kline JA, Courtney DM, Kabrhel C, Moore CL, Smithline HA, Plewa MC, Richman PB, O'Neil BJ, Nordenholz K. Prospective multicenter evaluation of the pulmonary embolism rule-out criteria. J Thromb Haemost. 2008;6(5):772-780. PMID: 18318689
3. Freund Y, Cachanado M, Aubry A, et al. Effect of the Pulmonary Embolism Rule-Out Criteria on Subsequent Thromboembolic Events Among Low-Risk Emergency Department Patients: The PROPER Randomized Clinical Trial. JAMA. 2018;319(6):559-566. PMID: 29450523
4. Liu X, Fan H, Hai Y, et al. Diagnostic Validation of PERC and Resource Utilization in Suspected Acute Pulmonary Embolism: A Systematic Review and Meta-Analysis. Clin Appl Thromb Hemost. 2026;32:10760296261472610. PMID: 42541393
5. Singh B, Parsaik AK, Agarwal D, Surana A, Mascarenhas SS, Chandra S. Diagnostic accuracy of pulmonary embolism rule-out criteria: a systematic review and meta-analysis. Ann Emerg Med. 2012;59(6):517-520. PMID: 22177109
6. Hugli O, Righini M, Le Gal G, Roy PM, Sanchez O, Verschuren F, Meyer G, Bounameaux H, Aujesky D. The pulmonary embolism rule-out criteria (PERC) rule does not safely exclude pulmonary embolism. J Thromb Haemost. 2011;9(2):300-304. PMID: 21091866