Executive summaryBackground  Vital signs traditionally consist of blood pressure, temperature, pulse rate and respiratory rate, and are an important component of monitoring the patient’s progress during hospitalisation. An initial search of the literature indicated that there was a vast volume of published information relating to this topic; however, there had been no previous attempt to systematically review this literature. This review was therefore initiated to identify, appraise and summarise the best available evidence relating to the measurement of vital signs in hospital patients.Objectives  The objectives of this review were to present the best available information related to the monitoring of patient vital signs with regard to their purpose, limitations, optimal frequency of measurements, and what measures should constitute vital signs. The review also sought to identify additional issues of importance related to the individual parameters of temperature measurement, blood pressure assessment, pulse rate measurement and respiratory rate measurement.Review methods  This review considered all studies that related to the objectives and included neonatal, paediatric and/or adult hospital patients. The outcome measures of interest were those related to the accuracy of, required frequency of or the need for vital signs. The review also considered any study addressing some aspect of vital signs measurement to ensure all issues of importance were identified. The search sought to find both published and unpublished studies. Databases searched included CINAHL, Medline, Current Contents, Cochrane Library, Embase and Dissertation Abstracts. The references of all identified studies were examined for additional references. All studies were checked for methodological quality, and data was extracted using a data extraction tool.Results  Although a variety of measures may be useful additions to the traditional four vital sign parameters, only pulse oximetry and smoking status have been shown to change patient care and outcomes. There are suggestions that vital sign monitoring has become a routine procedure, but little useful information was identified in regard to the optimal frequency of vital sign measurement. It was noted that many of the important issues related to vital sign measurement have not been investigated through research.There is currently only limited research related to respiratory rate as a vital sign; however, its value as an indicator of serious illness has not been reliably established. There is only limited research relating to pulse rate measurements. Although routinely used for all hospital patients, the ability to detect serious physiological changes by assessment of pulse rate has not been rigorously evaluated. Many factors were identified that could potentially influence the accuracy of blood pressure measurement. Auscultation is accurate for the measurement of systolic blood pressure using phase I Korotkoff sound as the reference point, and for diastolic pressure if phase V Korotkoff sounds are used. Cuff size can influence accuracy, in that using a cuff that is too narrow will likely overestimate blood pressure and a cuff that is too wide will underestimate the pressure. Research suggests that blood pressure should be measured on the upper arm, while the arm is resting at approximate heart level. Studies have shown that healthcare workers often measure blood pressure in an incorrect and inaccurate way, and this is of some concern. However, a small number of studies suggest that education programs can be effective in improving blood pressure measurement techniques. The largest volume of research identified during this review related to the measurement of temperature. For accurate measurement of oral temperatures the thermometer should be positioned in either the left or right posterior sublingual pocket and remain in the mouth for 6–7 min. Although oxygen therapy and different types of breathing patterns will not influence accuracy of oral temperature measurements, hot or cold liquids will. For the measurement of tympanic temperatures, an ear tug should be used to help straighten the external auditory canal and so ensure measurement accuracy. The presence of impacted cerumen will likely result in inaccurate measurements. The only potential harm as a result of measuring vital signs was associated with glass mercury thermometers, in terms of rectal perforation, the risk of mercury poisoning was not clearly established.Conclusions  Although there has been considerable research undertaken on many specific aspects of vital sign measurement, there is an urgent need for further primary research into the more general issues such as what parameters should be measured, the optimal frequency of measurements and the role of new technology in patient monitoring.