Orthostatic vital signs don’t help (White 2019)

Until I moved to New Zealand, I honestly thought that orthostatic vitals signs were dead. I never bothered to cover the evidence, because I thought we had all agreed they were pretty useless. However, in my current hospital, the trainees seem to routinely perform orthostatics, and the admitting medicine doctors frequently ask about them. In their defense, the current AHA guidelines recommend measuring orthostatic vital signs in the assessment of patients with syncope (although without much in the way of supporting evidence). (Shen 2017)

There has never been strong evidence to support orthostatic vital signs. If you test a random sample of asymptomatic, healthy, elderly patients, you will find abnormal results in a large number. (In other words, orthostatics aren’t very specific.) For example, Raiha (1995) looked at a random sample of well elderly patients, and found orthostatic changes in 28%, with no association with 10 year outcomes. Ooi (1997) looked at 911 asymptomatic nursing home patients, and found orthostatic hypotension in more than half. Similar results have been seen in adolescent patients. (Steart 2002) The sensitivity of orthostatics might be even worse. One systematic review reports that orthostatic vital signs have a sensitivity of 22% (95% CI 6-48%) for moderate blood loss. (McGee 1999) Prior emergency medicine research has also demonstrated no relationship between orthostatic vital signs and life threatening causes of syncope. (Schaffer 2018) Which brings us to the current study, an observational look at orthostatic vital signs in elderly patients with syncope…

The paper

White JL, Hollander JE, Chang AM, et al. Orthostatic vital signs do not predict 30 day serious outcomes in older emergency department patients with syncope: A multicenter observational study. The American journal of emergency medicine. 2019; PMID: 30928476 [article]

The Methods

This is a secondary analysis of a multi-centre prospective cohort study looking at elderly patients with syncope. (Clinicaltrials: https://clinicaltrials.gov/ct2/show/NCT01802398)

Patients

Patients 60 years or older presenting with syncope or near syncope. 

• Exclusions: intoxication; medical or electrical intervention to restore consciousness; a presumptive diagnosis of seizure, stroke, TIA, or hypoglycemia; orthostatic vitals not obtained or documents. 

Intervention

Orthostatic vital signs. 

• Abnormal orthostatic vital signs were defined as a systolic blood pressure drop of 20mmHg after two minutes of standing OR 10 mmHg upon standing OR symptoms of dizziness or lightheadedness upon standing.

Comparison

None

Outcome

The primary outcome was a large composite of 30 day serious events, which included a significant arrhythmia (ventricular fibrillation, symptomatic ventricular tachycardia, sick sinus syndrome, sinus pause more than 30 seconds, Mobitz II heart block, complete heart block, symptomatic supraventricular tachycardia, or symptomatic bradycardia less than 40 beats per minute), myocardial infarction, a cardiac intervention, new diagnosis of structural heart disease, stroke, pulmonary embolism, aortic dissection, subarachnoid hemorrhage, cardiopulmonary resuscitation, internal hemorrhage/anemia requiring transfusion, recurrent syncope/fall resulting in major traumatic injury, or death.

The Results

Out of the 6930 patients eligible, this trial included 1974 patients. Only half of patients included in the original trial had orthostatic vital signs measured. This population had a mean age of 72, with a relatively standard rate of comorbidities, and 50% of the included patients were female.

295 patients (15%) had a serious outcome by 30 days. The rate was the same whether or not the patient had abnormal orthostatic vital signs (14.7% vs 15.3%; OR 1.05; 95% CI 0.81–1.35; p=0.73).

Of the 20 different adverse events, there was no statistical difference for 18 of them. They report positive P values for GI bleed and stroke, but with so many comparisons, and no statistical adjustments made for multiple comparisons, these are likely due to chance alone. (I don’t think there is much biologic plausibility in orthostatics predicting stroke.) 

My thoughts

Although I think the results are correct, and fit with everything else we know about orthostatic vital signs, there are a number of limitations to this data. The biggest practical issue might simply be that the measurement of orthostatic vital signs is inconsistent, with measurement being taken in different positions and different times. Although the trial specifically defined abnormal, we can’t be sure how well the vitals were measured at each site.

Composite outcomes always worry me, and this composite included a very large variety of outcomes. On one hand, this is good, because we wouldn’t want to overlook any important adverse events. On the other hand, the large composite lumps events together which are clearly quite different, and could result is us overlooking an association between orthostatic vital signs and one specific diagnosis. Furthermore, powering the study for the composite means that many of the individual components happened very rarely, and the trial may have been underpowered to detect important differences in individual adverse events.

Selection bias is also a big issue, as there were a large number of exclusions in the original trial, and of those enrolled, only half had orthostatic vital signs measured and recorded. How this affects the results will depend a lot on why the patients didn’t receive orthostatic vitals. I think it could go both ways. Orthostatics are impossible to perform on really sick patients (who cannot stand), and there is no reason to perform them on patients you have already decided to admit to the hospital. On the other hand, physicians are also likely to skip orthostatics on really well patients. Realistically, the results are probably pretty representative of how orthostatics are used in the real world. 

In general, tests perform worse when applied indiscriminately. This study looked at the routine measurement of orthostatic vital signs and found no benefit. However, that doesn’t rule out a benefit in a subset of patients. Including a large number of patients in whom clinicians knew orthostatics were not going to be beneficial would only wash out a benefit, if it exists, in the smaller group where clinicians think they are important. So although this study tells us that we shouldn’t routinely be ordering orthostatic vital signs (and I think we already knew that), it doesn’t prove they are useless.

Bottom line

Despite the recommendation by the AHA, this is further evidence that routine orthostatic vital signs have no role in the evaluation of ED syncope patients.

(As I have discussed previously, guidelines that don’t discuss the quality of the evidence, omit the magnitude of the benefits, and mix evidence based recommendations with expert opinion are the bane of modern medicine.)

Other FOAMed

REBELEM: (Dis)utility of Orthostatics in Volume Depletion

REBEL EM: Still Worthless – Orthostatic Vitals and 30-Day Outcomes

---

Enjoying these emergency medicine EBM updates? Make sure to sign up, so you never miss a post:

---

References

McGee S, Abernethy WB, Simel DL. The rational clinical examination. Is this patient hypovolemic? JAMA. 1999; 281(11):1022-9. [pubmed]

Ooi WL, Barrett S, Hossain M, Kelley-Gagnon M, Lipsitz LA. Patterns of orthostatic blood pressure change and the clinical correlates in a frail, elderly population. JAMA 1997; 277: 1299-1304. PMID: 9109468

Raiha I, Luutonen S, Piha J, Seppanen A et al. Prevalence, predisposing factors and prognostic importance of postural hypotension. Arch Intern Med 1995; 155: 930-935. PMID: 7726701

Schaffer JT, Keim SM, Hunter BR, Kirschner JM, De Lorenzo RA. Do Orthostatic Vital Signs Have Utility in the Evaluation of Syncope? The Journal of Emergency Medicine. 2018; 55(6):780-787.

Shen W, Sheldon RS, Benditt DG, et al. 2017 ACC/AHA/HRS Guideline for the Evaluation and Management of Patients With Syncope Journal of the American College of Cardiology. 2017; 70(5):e39-e110.

Stewart JM. Transient orthostatic hypotension is common in adolescents. J Pediatr 2002; 140: 418-24. PMID: 12006955