Shared learning database

Portsmouth Hospitals NHS Trust
Published date:
February 2011

NICE CG50 requires that all adult patients in hospital have (i) a clear written monitoring plan specifying which vital signs should be recorded (and at what frequency), (ii) their severity of illness measured using a physiological early warning score (EWS) and (iii) a graded response strategy. VitalPAC, a wireless, handheld computer-based system, provides a structured mechanism for clinical staff to record patient vital signs electronically at the bedside on general hospital wards, calculates patients' early warning scores and provides instruction to bedside staff (e.g., 'increase the observation frequency', 'obtain assistance'). Importantly, all the data collected by VitalPAC is integrated with the hospital's other clinical systems and provides staff elsewhere in the hospital with instantaneous, reliable access to the charts and data via the intranet.

Does the example relate to a general implementation of all NICE guidance?
Does the example relate to a specific implementation of a specific piece of NICE guidance?


Aims and objectives

The aim was to reduce the number of avoidable adverse outcomes, such as cardiac arrest, unanticipated intensive care unit admission and unexpected death, in hospital. Preventable critical illness and death is receiving serious attention in developed healthcare systems, especially in acute hospital admissions. It is estimated that approximately 23,000 in-hospital cardiac arrests in the UK, and at least 20,000 unanticipated intensive care unit (ICU) admissions in England, Wales and Northern Ireland, may be avoidable with better care. Patients suffering these events often exhibit premonitory physiological deterioration. Abnormalities of pulse rate, blood pressure, temperature, etc., occur before 79% of in-hospital cardiac arrests and in 54% of in-hospital deaths and emergency ICU admissions. In some cases the deterioration is well documented, but with little discernable evidence of intervention. In others, the monitoring and recording of vital signs appears to be infrequent or incomplete. Our objectives included an improvement in the accuracy, availability, reliability and clinical impact of patients' vital signs datasets and EWS records. We believed that it was possible to develop a reliable, electronic method for collecting vital signs in a general ward area of the hospital and to establish a fully auditable trail of patient physiological data. This would provide opportunities to monitor the quality of care provided to acutely ill patients on general wards and to implement necessary change management, or educational processes, to improve patient outcomes. At the same time, it was hoped that we would improve compliance with agreed vital signs observation schedules, improve the application of the hospital's protocol for the urgent response to sick or deteriorating patients and reduce the nursing burden. A secondary aim was to produce a clinical dashboard (updated daily) permitting a hospital-wide view of activity in three categories - observations and associated workload (e.g., numbers of vital signs sets, completeness); compliance with protocols (e.g., late observations, mean delay, night-time observations); escalations and resources (e.g., risk category, time between observations).

Reasons for implementing your project

Portsmouth Hospitals NHS Trust was an early adopter of a paper-based early warning system. However, like many other hospitals, we had identified patient safety problems resulting from inadequate frequency of observations, incomplete observation sets, lack of knowledge of meaning of abnormal values, and a failure to call for assistance when required. Human factors also dictate that staff found it difficult to calculate early warning scores accurately and staff may also fail to inform each other regarding patients' care, particularly during handovers and transfers. Traditionally, a patient's vital signs are recorded on a chart held at the bottom of the patient's bed with consequent limits on the availability of these data to staff distant from the patient, e.g. critical care outreach team. All of the above lead to an inefficient system which creates significant risk to patients.

How did you implement the project

We believed that improving the processes associated with the detection, recognition and response to patient deterioration could best be improved by using an appropriate clinical IT system, which communicated directly with the caregiver; ensured processes and protocols are completed in the appropriate order; provided appropriate prompts where key pieces of information are required; performed necessary calculations accurately; provided decision support and produced legible documentation of clinical data. The greatest, challenge faced was to overcome the human resistance to change. For over 100 years, vital signs have been recorded on paper charts stored at the bottom of the bed. Now, vital signs are not only available at the bedside, they are accessible by staff anywhere in the hospital. Staff attitudes to the use of computers also varied from enthusiastic acceptance to outright hostility. Much of the resistance to change was overcome by including staff in the project development from the outset, ensuring that it fitted into, and enhanced, current nursing/medical working practices. Health Service IT projects have a history of failing to deliver on intended benefits. Further, many IT and other healthcare technology products fail to do what clinical staff required them to do. To avoid these pitfalls, we ensured that VitalPAC was a clinical project, which happened to involve computers, rather than an imposed IT solution. Many healthcare ICT failures are due to the IT system failing to fit comfortably with the clinical processes and the working patterns of doctors and nurses. ICT systems often seek to remodel clinical processes, whereas with VitalPAC, we sought to integrate IT with those processes. The hospital-wide implementation of VitalPAC costs ~£1 per patient bed day.

Key findings

Progress has been monitored using the real-time operational and clinical performance dashboards which provide the organisation with several instantaneous measures: workload, compliance with protocols and a risk profile of the organisation. We have also monitored our Trust's Hospital Standardised Mortality Ratio (HSMR) from before the start of the project until today. From the start of VitalPAC deployment in March 2005 to the present we have seen a fall in HSMR for the hospital of approximately 40%, this fall in HSMR increasing as VitalPAC usage spread. We have undertaken research which showed that VitalPAC facilitates the accurate calculation of early warning scores, thereby removing the error of human calculation. Our hospital VitalPAC team has developed the ViEWS early warning score, which is being used as the basis for a national score by the Royal College of Physicians, based on the unique, large vital signs database gathered using VitalPAC.

Key learning points

We feel that understanding how to improve the detection, recognition and response to patient deterioration requires the deconstruction of current processes to their component parts. By reconstructing the process in a fail safe way, inefficiencies can be eradicated and patient safety increased. Adopting a 'bottom up' approach to developing solutions ensures that new processes fit into, and enhance, current nursing/medical working practices. All staff members should be encouraged to provide critical appraisal throughout. We believe that it is essential to avoid imposed IT solutions as many IT and other healthcare technology products fail to do what clinical staff require of them.

Contact details

Professor Gary Smith
Consultant in Critical Care
Portsmouth Hospitals NHS Trust

Primary care
Is the example industry-sponsored in any way?

The submission describes a collaborative development between Portsmouth Hospitals NHS Trust and The Learning Clinic Ltd., London.