Making pain visible: an audit and review of documentation to improve the use of pain assessment by implementing pain as the 5th vital sign.
Liz Purser1 P.G. Cert
Kelly Warfield2 MSc
Cliff Richardson3 PhD
- Central Manchester University Hospitals Foundation Trust, UK
- High Wycombe Hospital, UK
- University of Manchester, UK
Correspondence address
Cliff Richardson
University of Manchester
School of Nursing, Midwifery and Social Work
Jean McFarlane Building
Oxford Road
Manchester
M13 9PL
Tel: +44(0)1613067639
Fax: +44(0)1613067707
Email:
Abstract
Background. Pain has been promoted as the 5th vital sign for a decade but there is little empirical evidence to suggest that doing so has affected the care of individuals suffering pain.
Method. This was a three stage audit of pain assessment in one large teaching hospital in the North West of England. Stage one measured the baseline pain assessment activity on surgical and medical wards and identified that the pain assessment tool was not visible to nurses. Stage two re-designed the patient observation charts held at the end of the bed and piloted two versions for clinical utility. Version 2 which had pain assessment alongside the early warning score was adopted and introduced throughout the hospital. Stage three audited pain assessment and management eight months after the introduction of the new charts.
Results. Pain was assessed more regularly at the stage three audit than at the baseline audit. On average pain was assessed alongside other routine observations 70% of the time across surgical and medical wards. Medical wards appeared to improve their pain assessment using the philosophy of pain being the 5th vital sign better than surgical wards as they assessed pain alongside routine observations in more than 90% of cases. Stage three identified that where a high pain score was recorded analgesia was delivered in the majority of cases (88%).
Conclusion. Introducing the philosophy of pain as the 5th vital sign and making pain assessment more visible on the patient observation chart improved the uptake of pain assessment. Pain management strategies were stimulated when high pain scores were identified.
Background
Several studies have shown that hospitalized patients experience inadequate pain relief and report moderate to high pain scores (Cousins et al., 1996; Powell et al., 2004; Shi et al., 2007; Sommer et al., 2008). This poor pain management incurs costs for the individual and to society. The individual costs include: psychological distress; reduced function; development of chronic pain; and reduced quality of life(Liu & Wu, 2007; Peters et al., 2010). Societal affects include: inability to work; and increased health care costs (Peters et al., 2010; VanDenKerkhof et al., 2006).
Pain assessment is a key component of successful pain management (Breivik et al, 2008).According to Breivik et al (2008) pain at rest and movement can be reliably assessed using tools such as the numerical rating scale or visual analogue scale, especially when looking for changes in pain intensity (Breivik et al 2000). One way that has been proposed to improve pain assessment is to regard pain as the 5th vital sign (JCAHO, 2001; Lanser & Gesell, 2001; Lynch, 2001a; Rousseau, 2008). The premise being that if pain was assessed with the same priority as the other vital signs, which are often carried out as part of the early warning score (EWS), unnecessary suffering and pain related complications could be avoided and patient outcomes improved (Lynch, 2001b; Merboth & Barnason, 2000).
Improving pain assessment and documentation has been shown to improve pain management (Morrison et al., 2006; Ravaud et al., 2004).Providing adequate pain management is a quality issue but studies designed to audit and research the effects of utilizing the principle of pain as the 5th vital sign have shown mixed results. A retrospective study using electronic medical records found no improvement in pain assessment (Mularski et al., 2006). However as the electronic device was not always available during assessments staff were not always prompted to record pain along with other vital signs hence these results should be viewed with caution. Ravaud and colleagues found greater numbers of patients having pain assessed following the introduction an education program which promoted pain as the 5th vital sign (Ravaud et al., 2004). It was however unclear as to whether there was improved pain management alongside the increased pain assessment.
Emphasis on the importance of measuring the impact of interventions on the quality of patient outcomes is urgently required to evaluate the effects of strategies such as making pain the 5th vital sign and whether implementing it makes a difference. This audit was proposed initially to identify if promoting pain assessment as the 5th vital sign affected nurses’ pain management behavior, however whilst undertaking a pre-audit review it became clear that this could not be introduced in isolation of making some other structural and contextual changes. A more complex change model wasneeded alongside the audit of the effects of the new philosophy for pain assessment.
Method
This was a three stage audit designed by the nursing pain team to identify current pain assessment practice and to monitor the effects of introducing the principles of pain as the 5th vital sign. As the audit stages progressed organizational learning occurred and led to the development of new observation charts and a short education program which assisted in implementing the new pain assessment philosophy. This learning led to adaptations of the final stage data collection due to gaps identified in the first stage by the data collectors.
Each stage revieweddocumentary evidence of pain assessment in one 24 hour period. The third stage reviewed analgesic administration alongside pain assessment in order to try to establish a link between assessment and management.
Stage one was an evaluation of current pain assessment practice. It was practice in our hospital to use the 0-10 numerical rating score to assess pain and the audit was employed to judge how widely the 0-10 score was being used. Every patient on eight surgical and five medical wards in the hospital were selected. All charts held at the end of the bed were reviewed by a member of the pain team to identify baseline levels of pain assessment. The review categorized the patients and analyzed how many had pain assessed. If pain was assessed a count of the number of times it was assessed was taken. No judgment was made as to the necessity or accuracy of the assessment. During this process however it was noted that the pain assessment chart was printed on the back of the observation chart held at the end of the patient’s bed. This was identified at this stage to be a potential inhibitor to the introduction of pain as the 5th vital sign which led to the development of stage two.
Stage two took account of the results from stage 1. It was a pilot of two new forms of observation chart. Both had pain assessment integrated alongside the other important patient observations rather than on the back of the chart. Version one (V1) required pain scores on movement to be plotted graphically along with temperature recordings and was piloted on one surgical ward. Version Two (V2) had boxes for recording pain at rest and on movement which were placed alongside the Early Warning Score (EWS). Qualitative feedback from the nurses who used the new observation charts was collected to find out their views about the chart usability and clinical utility.
Stage three occurred after the results of stage two were known. One of the new observation charts was preferred by the nurses who piloted it and was chosen to be introduced across the hospital. An education program written and delivered by the specialist pain nurses and targeted for all nurses was rolled out across the whole hospital. This included familiarization with the new observation chart which was only implemented once all education had been completed. Eight months after the introduction of the new chart into clinical practice a review similar to that performed at stage 1 was completed on the same wards. In order to judge the full extent to the change in practice additional factors were audited at stage 3. These included scrutiny of the new charts to establish whether pain was being assessed as the fifth vital sign. This was determined by identifying if pain was recorded at the same times as other observations. The second addition was designed to enable a measurement of the effectiveness of the pain assessment rather than just the fact that assessment was performed. Cross reference was made between the pain assessment and the delivery of analgesia by setting an arbitrary pain score of 5/10 or higher to represent this link. When a pain score of 5 or above was found the audit identified whether analgesia was given and then if the pain score reduced below 5 within an hour of the first recording.
Results
Stage One
All of the selected thirteen ward areas participated however the number of forms returned by each ward varied between four and twenty-four. A total of 212 forms were returned but ten were only partially completed and were discarded leaving a total of 202 patients included in the analysis. Nearly half of the patients were post-surgical (88; 44%) or non-surgical (99; 49%) with a small group of pre-surgical patients (15; 7%). Twelve patients (6%) were identified to have a chronic pain condition either as the reason for their admission or additional to their reason for admission (i.e. admission for surgery but had an underlying chronic pain condition). The average age of the patients from each ward varied considerably with one ward’s average age being 28 years (12 patients) and another being as high as 74 years (17 patients). The overall average age was 54 years.
A large majority (171; 85%) of the patients did not have pain assessment recorded on their charts. Of the 31 (15%) patients that had pain assessed, 20 had it assessed more than once with the largest number of assessments in any one patient being 12. The average number of assessments performed in the patients who had pain assessed was 3.4 but the mode was 1.
Stage Two
Following the results of stage 1 it was considered important to try to find new ways to assist nurses in assessing pain. One way to do this was thought to be to move the pain assessment tool from the back of the chart to the front to make it more visible. Two pilot versions of a new patient observation chart were devised and tested.
Version 1 (V1) was tested on 23 patients on one surgical ward. Version 2 (V2) was piloted on 37 patients on one surgical and one medical ward. Despite the fact that the pain score was visible on both versions of the chart,pain assessment was performed on 100% of the patients on which V2 was used but only on 34% of patients where V1 was trialed.
Qualitative data collected from the nurses using the pilot versions resoundingly supported the use of V2. Examples of the positive comments were:
‘…chart’s excellent’
‘…would like to continue using this.’
‘…good that it is close to the EWS.’
Whereas comments associated with V1 were generally negative. Examples of comments written about V1 were:
‘…confusing and messy.’
‘…would be better as numbers.’
‘…would be better at the bottom as part of the EWS.’
As Version 2 was identified as having the better clinical utility it was adopted as the new observation chart and was introduced across the hospital.
Stage Three
The same eight surgical wards and five medical wards used in stage 1 took part in stage three. A total of 253 patients data were analyzed however some data sheets were not fully completed hence the denominator in the results varies.
The type of patient was similar to stage one (see table 1) as was the average age of 58 years.
Table 1. Comparison of patients audited at stage 1 and stage 3.
Thirty four patients (14%) were identified to have a chronic pain condition. These were found on a mixture of surgical and medical wards. Eighteen (7%) patients had patient controlled analgesia (PCA) or epidural analgesia at the time of the audit.
Pain assessment
Ninety-six percent of patients had at least one pain score documented. This compared with 15% in stage 1. The frequency of assessment also improved between audits with 83% of patients having 3 or more pain scores documented in stage three (see Figure 1).
Figure 1 Percentage of patients who had pain assessed in stage 1 and stage 3.
Fifth vital sign
Pain was always recorded alongside the other vital signs in 70% of patients. Differentiating by ward type it was found that medical wards linked the recording of these observations more often than the surgical wards (See Figure 2).
Figure 2. Percentage of patients where pain was assessed alongside the recording of other observations.
The effect of pain assessment
Sixty five patients (26%; n=253) had at least one pain score of 5/10 or higher during the audit period and 57 (88%) of these had a pain management intervention immediately or close to the assessment time. Most of the interventions were to administer already prescribed analgesia (65%) but in a small amount of cases the analgesia was altered (17%) or an additional bolus dose of analgesic was prescribed and delivered (6%) (see table 2). Seven of the eight additional bolus doses were given on medical wards so were not thought to be linked to PCA as these are rarely used on medical wards in our hospital and could not be associated with epidural infusions as these are not used on general wards.
Table 2. Interventions made for a pain score greater than 5
Following the intervention a pain score was recorded within an hour in 29 patients (44%) and in the majority (56%) the reassessment score was less than 5/10. Of the sixty five patients with a score of 5/10 or higher, 15 (26%) were those identified to have chronic pain and hence may not be expected to respond in the same way as those with acute pain.
Discussion
Audit and practice development can go hand in hand. A new initiative to promote pain as the fifth vital sign was considered, but in order to judge its effectiveness, data were required from the time before and after the introduction of the initiative. The stage 1 audit identified that pain assessment within the hospital was rarely performed, but at this time it was thought that one potential reason for this lack of assessment was the visibility of the pain assessment tool. All pain assessment information was moved to the front of the patient observation chart.
The pilot of two newobservation charts identified that visibility in itself was not sufficient to improve nursing adherence to pain assessment. A pilot chart, where the pain assessment scale was both visible and placed next to the Early Warning Score (EWS) achieved higher levels of adherence compared to one where pain assessment was visible and plotted alongside the patients’ temperature.Nurse’s qualitative responses suggested that the chart with pain assessment alongside the EWS also had a high level of clinical utility. Piggybacking the recording of pain assessment to the prominent EWS magnified the effect of the improved visibility.
Additionally a short education program targeted towards the use of pain as the fifth vital sign was delivered to nurses across the hospital. This was in the form of a powerpoint presentation which was placed on the desk top of all ward computers. All nursing staff was asked to read through it,andonce completed, sign a ward list of names to enable the pain team to identify the level of compliance.
There were some results from the stage 3 audit which are notable. Firstly although more surgical wards were audited than medical wards the number of patients using the more technical forms of pain relief such as PCA and epidural infusions was low. Patients with PCA and epidurals often have specialized observation charts including pain assessment and compliance with these charts has been high (Rawal, 1999). The majority of patients in this sample were therefore the hard to reach group of patients where traditionally the recording of pain assessment has generally been low(Rawal, 1999). Improving the recording of pain assessment in this group is welcomed. Patients on medical wards had pain assessed alongside other observations in more than 90% of cases which supports the philosophy of making pain the fifth vital sign. Such a good result supports the fact that making pain assessment visible alongside a short education program can affect nursing practice. It could also however be argued that linking pain assessment to routine observation measurement could reduce the individualization of nursing care and potentially restrict nurses to only assessing pain at these times rather than when the patient requires it.
As pain assessment was performed regularly on the majority of patients at stage 3 it was possible to explore if performing pain assessment had an effect on pain management. Pain management interventions were not measured at stage 1 so it was not possible to compare between stage 1 and stage 3 however the vast majority of patients (88%) with high pain scores at stage 3 had analgesia delivered. Most were given already prescribed analgesics but some had newly prescribed analgesics including new bolus (one off/stat) doses which may signify that nurses were using more complex decision-making judgments. Significantly once delivered the analgesics reduced the pain below the threshold chosen for this audit which was 5/10, in more than half (56%) of the cases. It is appreciated that the likelihood of relief in chronic pain patients using analgesia in this way is less than it is for acute pain. Removing the chronic pain patients from this analysis may well have increased the numbers that successfully reduced below 5/10 however due to the nature of the way data were collected it was impossible to identify the acute pain patients from the chronic pain patients. Future audits will be modified to enable capture of this important aspect of pain management. It is recognized that the cutoff point (5/10) for intervention was arbitrary however it was chosen to represent the effects of pain assessment on pain management and should not be evaluated as a statement to the effect that only scores above 5 should receive relief.