Abstract
We conducted a retrospective review of the digital images in the clinical records of 50 patients with pressure ulcers. Ten nurses independently assessed one image from each patient. There was a total of 414 responses from the nurses about the stage and location of the ulcers (83% response rate). The average agreement about the stage and location of the ulcers was 85%. The overall agreement declined as the stage of the ulcer increased. The average agreement regarding the wound descriptors was: necrosis 85%, granulation tissue 81%, ischaemia 83%, cellulitis /infection 69%, erythema 68%. Almost all nurses felt the need to change the current management of the wound (460 responses, or 92%). The nurses judged that most digital images were of good quality (17%) or very good quality (79%). The present study suggests that a high percentage of assessments for patients currently travelling to specialist clinics could be performed in the community using digital images and telemedicine.
Introduction
People with spinal cord injury (SCI) who have completed their initial rehabilitation remain at life-long risk of developing pressure ulcers. Considerable time is spent in educating them about the self-care necessary to prevent this complication. Pressure ulcers are the commonest medical problem in chronic SCI patients. 1 The annual incidence is 23–30% and the lifetime risk is over 70%. 2
Telemedicine may be a useful technique in the prevention and management of pressure ulcers in patients with SCI. Two studies have been carried out to determine whether wound care via videoconferencing was a satisfactory alternative to clinic visits.
3,4
In another study, it was found that telerehabilitation allows a nurse to assess a patient's pressure ulcer, consult appropriate caregivers and make recommendations for treatment.
5
These results are all encouraging. However, there do not appear to have been any published studies about using digital images to assess pressure ulcers in patients with spinal cord injury in the UK. We have examined the feasibility of using digital images for pressure ulcer assessment and for reviewing the management options. The following questions were addressed:
Was the quality of the image satisfactory for making wound management decisions? Does the wound management need to be changed?
Methods
We conducted a retrospective review of the clinical records of 50 patients referred with pressure ulcers to the National Spinal Injury Centre (NSIC) outpatient department over a 12-month period. The patients studied all had a spinal cord injury, had received rehabilitation at the NSIC, had a pressure ulcer and were aged 18 years or over. The presence of a pressure ulcer had previously been diagnosed by trained nursing staff and a full pressure ulcer assessment form had been completed.
The study was designed to compare assessments made from digital images with the true or reference values that had been recorded previously in the clinical notes. Digital images of pressure ulcers were already in the case notes, having been taken as a routine part of pressure ulcer assessment in the outpatient department.
Image assessment
The digital images were assessed by nursing staff using assessment forms based on the European Pressure Ulcer Advisory Panel framework. The variables assessed were pressure ulcer location, stage of pressure ulcer, state of the pressure ulcer bed (slough, necrosis, granulation, epithelization), characteristics of the pressure ulcer edge (maceration, oedema, induration) and surrounding skin condition (eczema, inflammations/infection). The four TIME variables were also assessed: Tissue, Infection or Inflammation, Moisture imbalance and Edges non-advancing. 6 The nurses were also asked about the quality of the images and whether this influenced their wound management decisions. Ten nurses independently assessed all 50 images.
After the assessments had been completed, the nurses were asked to identify barriers to the use of digital photography in clinical practice. They chose from the following list:
Cost and maintenance of equipment; Resistance to technology; Competence in taking good quality photographs; Confidentiality and privacy; Accuracy and reliability.
Reference values
The true or reference variables were taken from the case notes and related documentation. The reference values were recorded at the same time that the digital image was taken. The reference values were recorded by nurses different from the ones who reviewed the digital images. An example of the variables recorded is shown in Table 1.
Example of estimated and reference values
Results
The mean age of the patients was 69 years (range 30–90) (see Table 2).
Characteristics of the 50 patients
*American Spinal Injuries Association classification of spinal cord injury
Agreement
There was a total of 414 responses from the nurses about the stage and location of the ulcers (83% response rate). The average agreement about the stage and location of the ulcers was 85%. The highest agreement was for stages 1 and 2 (100%) and the lowest for stage 4 (67–77%) (see Table 3). The overall agreement declined as the stage of the ulcer increased (see Figure 1).
Agreement between nurses about wound stage
Agreement about stage and location of ulcer
The average agreement regarding the wound descriptors was: necrosis 85%, granulation tissue 81%, ischaemia 83%, cellulitis/infection 69%, erythema 68%.
Wound status
The nurses were able to provide an adequate written description of the wound. The status of the ulcers is summarised in Table 4. Almost all nurses felt the need to change the current management of the wound (460 responses, or 92%) (see Table 5).
Evaluation of pressure ulcer status. The values shown are the number of times (%) that ten nurses assessed the characteristic as being present in the 50 digital images
Responses from nurses who indicated that the wound management needed to be changed
Nurses' opinions
The nurses judged that most digital images were of good quality (17%) or very good quality (79%). Only 4% of images were judged to be of insufficient quality.
The nurses commented on five barriers to incorporating digital photography into routine wound care assessment in the spinal outpatient department (Table 6).
Barriers to digital photography identified by the nurses (n = 10)
Discussion
In assessing wound healing, it has been argued that clinicians are accustomed to making decisions with relatively incomplete information. 7 Relative to face-to-face consultation, there is a reduction in the available information with the use of telemedicine, e.g. odour in wounds cannot be assessed remotely. Jones et al. also stated that to be effective and safe, the process needs to demonstrate an acceptable level of accuracy and reliability. Accuracy is reflected by the degree of concordance between the telemedical and face-to-face diagnosis. Reliability depends on how consistently a set of results is reproduced with different operators. In the present study the nurses were able to give accurate and reliable assessments of the ulcer bed, ulcer edge and the condition of the surrounding skin. This shows that digital images are satisfactory for wound assessment.
The findings of the present study agree with recent research regarding the known location of pressure ulcers in patients with spinal cord injury. 8 We also observed that the complexity of the pressure ulcer stage appeared to affect the agreement rate among nursing staff making the assessments (see Figure 1). That is, there was a lower concordance rate as the pressure ulcers became deeper and more complex. Further research is needed to clarify the structure and process changes that are required to make best use of digital images for assessing pressure ulcers in spinal cord injury patients.
The present study suggests that a high percentage of assessments for patients currently travelling to specialist clinics could be performed in the community using digital images and telemedicine.
Footnotes
Acknowledgements
We thank the staff at the outpatient department at the National Spinal Injuries Centre at Stoke Mandeville Hospital for assisting with data collection.