3 Clinical evidence

Summary of clinical evidence

3.1 The key outcome for moorLDI2-BI laser doppler blood flow imaging is the development of an appropriate treatment plan based on an accurate assessment of burn wound depth and healing potential. Other performance measures are the sensitivity, specificity, negative predictive value and positive predictive value of the wound healing potential before 14 or 21 days. Clinical utility outcomes associated with the technology are avoiding unnecessary operations, extent of surgery, number of dressing changes, complications and length of stay in hospital. Longer-term outcomes are extent and type of scarring and the recovery of pre-injury function.

Accuracy of the moorLDI2-BI

3.2 The accuracy of the moorLDI2-BI in the assessment of burn wounds was examined in eight studies with a variety of criteria including ability to predict healing within 14 or 21 days. Comparisons were made with clinical and histological evaluation of burn wound depth.

3.3 Pape et al. (2001) reported an audit of wound healing at 21 days for 76 intermediate depth wounds in 48 patients. Results showed the moorLDI2-BI to be 97% accurate (74/76) in predicting wound healing at 21 days compared with 70% (53/76) for clinical evaluation (no statistical comparison reported).

3.4 Hoeksema et al. (2009) investigated the changing accuracies of laser doppler imaging and clinical evaluation over days 0, 1, 3, 5 and 8 after injury. Forty patients with intermediate depth burn wounds were scanned using the moorLDI2-BI. The final assessment of wound depth showed a deep partial or full thickness burn in 14 patients, 12 of whom had a skin graft, and a superficial dermal burn in 26 patients. Accuracies on days 0, 1, 3, 5 and 8 were 41%, 62%, 53%, 71% and 100% respectively by clinical evaluation, and 55%, 80%, 95%, 97% and 100% respectively by laser doppler imaging. The burn wound depth accuracy using the moorLDI2-BI was significantly higher than clinical evaluation on day 3 (p < 0.001) and day 5 (p = 0.005) but not on days 0, 1 or 8.

3.5 Jeng et al. (2003) described a prospective blinded trial comparing laser doppler imaging using the moorLDI2-BI versus clinical evaluation by an experienced burn wound surgeon to decide whether or not to operate. Forty-one wounds of intermediate depth were analysed. Biopsy confirmation was obtained for 21 wounds. There was agreement on wound depth between laser doppler imaging and clinical evaluation in 56% (23/41) of cases. The surgeon's determination of burn wound depth was accurate in 71% (15/21) of wounds biopsied. The moorLDI2-BI was 100% (7/7) accurate in wounds for which it indicated a need for excision.

3.6 Monstrey et al. (2011) compared healing prediction based on interpretation of a moorLDI2-BI scan with actual wound healing as recorded photographically for 433 burn wounds in 139 patients. This assessment found an overall accuracy for the moorLDI2-BI of 96.3% with sensitivity 94.5%, specificity 97.2%, positive predictive value 94.5% and negative predictive value 97.2%.

3.7 La Hei et al. (2006) scanned 50 burns in 31 paediatric patients. Two experienced burn wound surgeons independently reviewed the scans, photographs and a basic patient history, without meeting the patient. One surgeon identified 82 areas of differing depth, the other identified 76 areas, and both surgeons predicted healing times (superficial heal: less than 14 days or deep heal: more than 14 days or graft). Overall, 97% (154/158) of predicted healing times were correct with four deep burn areas incorrectly predicted to heal within 14 days. No superficial wounds were reported as deep.

3.8 Holland et al. (2002) investigated the ability of laser doppler imaging to evaluate burn wound depth in children by scanning 58 patients and comparing the predicted outcome (from either the scan or from clinical evaluation) with the subsequent wound outcome at 12 days. One patient was excluded because there was too much movement for the scan to be interpreted. Clinical evaluation correctly identified 66% (19/29) of deep partial or full thickness burns between 36 and 72 hours after injury compared with 90% (26/29) using moorLDI2-BI scans. Scans using moorLDI2-BI were also more specific, correctly diagnosing 96% (27/28) of superficial partial thickness burns compared with 71% (20/28) from clinical evaluation alone (no statistical comparison reported).

3.9 Niazi et al. (1993) reported results from a pilot study that analysed 17 burn wounds on 13 patients. Punch biopsies were used to confirm burn wound depth at 72 hours after injury. Clinical evaluation was correct for 41% (7/17) of the burns, overestimated depth in 41% (7/17) and underestimated depth in 18% (3/17). Burn wound depth assessed from moorLDI2-BI scans was correct for 100% (17/17) of burn wounds (no statistical comparison reported).

3.10 Mill et al. (2009) compared moorLDI2-BI image colours with wound outcomes in 85 burns on 48 children. Analysis of the image colour regions was found to be significantly related to re-epithelialisation (p < 0.003), grafting (p < 0.001) and active scar management (p = 0.003).

Clinical utility outcomes

3.11 Two studies evaluated whether or not using the moorLDI2-BI enabled appropriate skin grafting decisions to be made earlier than using clinical evaluation alone. Jeng et al. (2003) described a prospective blinded trial that compared laser doppler imaging versus clinical evaluation by an experienced burn wound surgeon, in deciding whether to operate or not on 41 burn wounds of intermediate depth. There was agreement on wound depth between the imaging and clinical evaluation in 56% (23/41) of cases. In these cases the moorLDI2-BI determined wound depth a median of 2 days (minimum = 0, maximum = 4) earlier than clinical evaluation alone (no statistical comparison reported). Kim et al. (2010) described a non-randomised cohort study of 196 children with an acute burn injury who required surgical treatment. Laser doppler imaging was used in addition to clinical evaluation on 49% (96/196), and 51% (100/196) were assessed by clinical evaluation alone. The mean time from date of injury to the decision to graft was 8.9 days in the moorLDI2-BI group compared with 11.6 days in the group assessed by clinical evaluation alone (p = 0.01).

Committee considerations

3.12 The Committee considered that there was good clinical evidence that information from moorLDI2-BI scans increases the accuracy of predicting burn wound healing and also that this information can be used to facilitate treatment plans. Using the moorLDI2-BI in addition to clinical evaluation can enable earlier surgical treatment in some patients and avoid the need for surgery in others. It may also reduce the extent of surgery.

3.13 Burn wounds on dark skin can be difficult to assess clinically. The Committee considered that the moorLDI2-BI offers particular advantages for assessing burn wounds on dark skin.

3.14 The Committee was advised that additional patient and system benefit could be obtained by using the moorLDI2-BI to define accurately the margins of surgical areas for the skin graft operations, so helping to limit the extent of excision and grafting in some patients.

3.15 There are many factors that are known to have a detrimental effect on moorLDI2-BI images or their interpretation, including infected wounds, patient movement, old scars and tattoos. These are acknowledged in the published studies and also recognised by the manufacturer in its user guide. It was therefore considered important that moorLDI2-BI images should only be taken and interpreted by a clinician trained in use of the technique.

3.16 The Committee was advised that the moorLDI2-BI can be used to assess burns treated by biological and semi-biological dressings.

3.17 The Committee considered that there was no evidence to suggest patients were likely to be harmed by the moorLDI2-BI used by trained clinicians.

  • National Institute for Health and Care Excellence (NICE)