Routinely Measured Vital Signs on General Hospital Wards, a Sacred Cow or Cardinal?

Monday, 30 July 2012: 3:05 PM

Marja N. Storm-Versloot, MSc
Department of Surgery, G3-151, Academic Medical Center, Amsterdam, Netherlands
Lotte Verweij, MSc
Department of Neurosurgery, Academic Medical Center, Amsterdam, Netherlands
Cees Lucas, PhD
Department of Clinical Epidemiology, Biostatistics & Bioinformatics, Academic Medical Center, Amsterdam, Netherlands
Jeroen Ludikhuize, MD
Department of Quality and Process Innovation, Academic Medical Center, Amsterdam, Netherlands
Dink A. Legemate, MD, PhD
Department Vascular surgery, Academic Medical Center, Amsterdam, Netherlands
J. Carel Goslings, MD, PhD
Department Trauma surgery, Academic Medical Center, Amsterdam, Netherlands
Hester Vermeulen, RN, PhD
Department of Quality Assurance & Process Innovation, Academic Medical Center at the University of Amsterdam, Amsterdam, Netherlands

Learning Objective 1: The learner will be able to acknowledge that diagnostic accuracy research is warranted on the individual contribution of single vital signs.

Learning Objective 2: The learner understands that commonly performed nursing practices can be questioned as they can be based on intuition or tradition, instead of scientific evidence.


As a daily routine hospital nurses perform numerous vital sign measurements. However this ‘sacred cow’ is questioned as evidence is conflicting on the clinical relevance in relation to adverse events.

Objective: To determine the clinical relevance of routinely measured vital signs (temperature, heart rate, blood pressure, oxygen saturation and respiratory rate) in relation to mortality, septic or circular shock, ICU admittance, bleeding, re-surgery or infection in ‘not at risk’ medical and surgical hospitalized patients.

Methods: For this systematic review we searched MEDLINE, EMBASE, CENTRAL, CINAHL, and MEDION to October 2011. Prospective studies evaluating the clinical relevance of vital signs in relation to adverse events were selected. Using structured forms, data and area under curve (AUC) was extracted independently by two authors from each study. If possible, 2*2 tables were constructed for each vital sign to recalculate predictive values and likelihood ratios.  

Results: Of 13.806 citations, one diagnostic accuracy study and 14 observational studies met the inclusion criteria. All were published between 1986 and 2010, with a total of 6019 participants. Eight studies concerned multiple vital sign measurements and seven only one. Due to heterogeneity, data was not pooled. Although some discriminative positive likelihood ratios were found, this was mostly when vital signs deviates strongly from normal values, e.g. respiratory rate <6 or ≥35/min, systolic blood pressure <90 mmHg, and oxygen saturation <85%. The accompanied positive predictive value was low to moderate. The highest AUC was 76% when three vital signs deviate.

Conclusion: Scarce evidence is found on the clinical relevance of routine vital sign measurement and no evidence based recommendation could be distilled from the current literature. There is a need for further prospective clinical relevance research on single or combination of routine vital sign measurement. Daily routine practices and nurses clinical decision making should be supported with evidence.