Prehospital vital signs can predict pain severity: analysis using ordinal logistic regression

Abstract

Objective 

To determine whether vital signs in patients suffering from acute pain in the out-of-hospital setting have any association with pain severity measured using an ordinal pain scale.

Materials and methods 

We conducted a retrospective analysis of over 53 000 adult patients aged between 16 and 100 years, who presented to paramedics complaining of acute pain between 1 January 2004 and 30 November 2006. Simple correlation (Spearman's) and ordinal logistic regression techniques were used to create a proportional odds model to explore the relationship between patient-reported pain score and initial vital signs including respiratory rate, pulse rate and blood pressure.

Results 

There was a weak but significant correlation between respiratory rate and initial pain score (R=0.15, P<0.0001). In patients aged 16 years and above, those with an initial respiratory rate of 25 breaths/min or more had significantly increased odds (45–105%) of having more severe pain than patients with a respiratory rate of less than 25 breaths/min (P<0.0001). In younger patients (aged between 16 and 64 years), a heart rate of 100 beats/min or more was associated with 18% increased odds of more severe pain (P<0.0001). In older patients (aged between 65 and 100 years), systolic blood pressure of 140 mmHg or more was associated with 14% increased odds of more severe pain (P<0.0001).

Conclusion 

An association between prehospital vital signs and pain severity has been shown using ordinal logistic regression. In adults, a respiratory rate of 25 breaths/min or more was the most important predictor of having more severe pain. Tachycardia and systolic hypertension may also be important in younger and older patients, respectively. Simple correlation fails to show clinically important associations between prehospital vital signs and pain severity.