UPPER/LOWER assessment
A total of 43 wound care patients participated in this observational study (1 wound per patient). Of the participants, 62.8% (27/43) were male, and age of patients ranged from 22 years old to 98 years old (Table 1). Wounds of all etiologies were assessed, including diabetic foot ulcers (9), pressure injuries (12), venous leg ulcers (VLUs) (8), surgical site infections (5), and other skin injuries and ulcers (9). The UPPER/LOWER wound assessment criteria were employed by clinicians at initial evaluation of these wounds. Presence of 3 or more signs and symptoms from each checklist (UPPER or LOWER) denoted an infection. Of the wounds assessed, 60.5% (26/43) as 61% in this case were positive for UPPER signs and symptoms of infection, indicating local infection. In contrast, 9.3% (4/43) of wounds were positive for LOWER signs and symptoms of infection, denoting deeper bacterial infection.
Microbiology swabs were collected from 27 of the wounds assessed; 81.5% of wounds sampled (22/27) were positive for bacteria, and of those positive for bacteria, 77.3% (17/22) had a heavy growth of bacteria. A diverse variety of Gram-positive and Gram-negative aerobic and anaerobic bacterial species were detected from these chronic wounds. The most prevalent species detected included: Staphylococcus aureus, Pseudomonas aeruginosa, Klebsiella, Staphylococcus lugdunensis, Escherichia coli, Enterococcus faecium, mixed anaerobes, and others. Microbiological results were compared with the UPPER/LOWER criteria to determine the sensitivity and specificity of these signs and symptoms in predicting wound infection. Of the wounds sampled, 18 were positive for infection according to UPPER/LOWER criteria and microbiological analysis confirmed presence of bacteria in these wounds. There were 5 true negatives (ie, negative for UPPER/LOWER and microbiology); no false positives were observed. The UPPER/LOWER criteria produced a diagnostic sensitivity of 81.8%, specificity of 100%, and PPV and negative predictive values (NPV) of 100% and 55.5%, respectively. Accuracy of the UPPER/LOWER checklists was 85.2% (Table 2).
Bacterial fluorescence imaging
Following evaluation of UPPER/LOWER criteria, standard and fluorescence images were captured for all wounds. Of the 43 wounds assessed, 38 (88.3%) exhibited bacterial (red or cyan) fluorescence, while 5 were negative for bacterial fluorescence. Examples of standard and fluorescence images captured, and their corresponding microbiology, are depicted in Figure 2. Addition of fluorescence information to the LOWER checklist criteria resulted in 10 wounds (23.3%) having 3 or more criteria of LOWER infection. The addition of fluorescence information resulted in 8 wounds (18.6%) fulfilling the criteria for UPPER infection. In total, fluorescence imaging helped to confirm superficial or deep infection in 62.8% (27/43) of wounds.
The utility of FL information at time of assessment was evaluated in 2 ways: (1) FL information was either added as an additional “check” to the UPPER and LOWER criteria or (2) it was used as an independent predictor of high bacterial loads indicative of potential wound infection (Table 2). A Fisher's exact test was used to assess statistical difference in diagnostic accuracy between UPPER/LOWER alone and with addition of FL. Addition of FL as an additional check in the UPPER/LOWER checklists significantly improved diagnosis of infected wounds compared with UPPER/LOWER criteria alone (P < .01). Sensitivity was increased by 13.6% up to 95.4%, and NPV and accuracy also were improved (83.3% and 96.3%, respectively). When FL information was used independent of UPPER/LOWER criteria to identify wounds with bacterial loads greater than 104 CUF/g, 11 additional wounds were identified as positive for bacterial loads greater than 104 CFU/g compared with UPPER/LOWER criteria alone. This resulted in 22 true positives, 5 true negatives, and no false negatives or false positives corresponding to sensitivity, specificity, NPV, PPV, and accuracy of 100%, per wound microbiology data. In all cases where cyan fluorescence was sampled, microbiology confirmed the presence of Pseudomonas aeruginosa. Due to the low number of true negative wounds (n = 5), specificity and NPV should be interpreted with caution.
Fluorescence-guided treatment outcomes
The information obtained from fluorescence images complemented the signs and symptoms reviewed as part of the UPPER/LOWER checklist assessment and was also used to guide treatment decisions. Fluorescence images greatly impacted treatment by facilitating more thorough cleansing and debridement in most study wounds. Fluorescence images also guided antimicrobial stewardship decisions, such as leading to avoidance of antibiotic administration (n=2), adding antibiotics where needed (intravenous [IV] = 2, oral = 1), and guiding the selection of appropriate antimicrobial dressings. In 3 instances, fluorescence information resulted in more frequent dressing changes. The impact that FL had on influencing treatment decision-making when combined with UPPER/LOWER checklist assessment is highlighted in the following case studies. Perhaps most importantly, fluorescence images improved communication to other members of the care team, resulting in clinically necessary, expedited interventions in 13.9% (6/43) of cases.
Case 1
In this patient with a DFU, local/superficial infection was diagnosed based on the presence of 3 UPPER symptoms: pain, exudate, and reek. Fluorescence imaging revealed red fluorescence indicative of bacteria at loads greater than 104 CFU/g in the wound (Figure 3A), supporting the diagnosis of local/superficial infection. The substantial degree of red fluorescence observed in the wound promoted the infectious disease physician to prescribe oral antibiotics.
Case 2
A patient developed an abdominal midline dehiscence wound (Figure 3B) after Hartman’s surgical procedure for a perforated bowel. The patient was previously on IV antibiotics and was about to be discharged. Based on the presence of 3 UPPER symptoms (unhealthy wound, exudate, and reek), the patient was diagnosed with a local/superficial infection. Bacterial (red) fluorescence was observed in the wound and confirmed the diagnosis of the local infection. The considerable amount of red fluorescence observed in and around the wound prompted the clinician to prescribe 2 oral antibiotics and request follow-up within 5 days. Swabs of the wound and microbiological analysis later confirmed heavy growth of K pneumoniae, S lugdunensis and mixed anaerobes.
Case 3
A patient with a venous leg ulcer exhibited edema and redness suggesting potential for deep infection according to the LOWER checklist. The wound had no observable signs and symptoms of Pseudomonas aeruginosa infection, but FL revealed significant cyan, glowing fluorescence signature (indicative of the presence of Pseudomonas) in and around the wound (Figure 3C). The presence of this fluorescence signature added an additional check to the LOWER checklist, confirming a deeper infection. The real-time information provided by the fluorescence images were used to target cleaning of the wound and a Pseudomonas-targeted dressing was selected. A notable decrease in cyan/white fluorescence signature was observed after cleansing with modified sodium hypochlorite solution.
Case 4
A patient with a pressure injury presented with a healing dehisced flap procedure. Using the UPPER/LOWER checklist to guide assessment, exudate and reek were observed, but no LOWER symptoms were present. Fluorescence imaging at the bedside revealed bacterial (red) fluorescence in the wound and brought the total number of checks for the UPPER assessment to 3, indicating the presence of surface contamination. The fluorescence images provided information on where to clean the wound; after surface scrubbing with a modified sodium hypochlorite solution, no red fluorescence remained (Figure 4A). A swab of the wound post-cleansing further confirmed the presence of surface contamination as only light growth of mixed coliforms, and beta hemolytic Streptococcus was observed after eradication of the red fluorescent signal.
Case 5
A patient with a pressure injury and obvious signs of superficial infection underwent assessment for deep infection and FL. Only 2 symptoms of LOWER were observed, but FL revealed presence of significant bacterial (red) fluorescence in and around the wound (Figure 4B), which indicated the presence of deep infection. Culture results confirmed heavy growth of methicillin-resistant S aureus. The wound was cleansed, resulting in a reduction in observable bacterial fluorescence. The substantial amount of red fluorescence observed prompted expedited intervention. A computed tomography (CT) scan was done, revealing the presence of an abscess. The clinician notified the plastic surgeon and plans ensued for scheduling an operating room procedure for incision and drainage, including a washout of the affected area.
Case 6
A 77-year-old patient with a dehisced abdominal incision postoperatively undergoing negative pressure wound therapy (V.A.C. Therapy; 3M + KCI) to help heal the wound. Two weeks after surgery, assessment of the wound revealed signs of UPPER infection (all criteria met) in addition to 3 LOWER symptoms (larger wound, warmth, and redness). Fluorescence imaging revealed bright red fluorescence in the wound bed (Figure 5). Swab analysis of the wound indicated heavy growth of mixed anaerobes and light growth of mixed coliforms. Information on bacterial load and location through FL prompted targeted cleansing and debridement of red (bacterial) fluorescing regions of the wound using saline and an antimicrobial wound cleanser. The bright red fluorescence that was observed in FL images triggered the wound care clinician to initiate IV antibiotics and request an expedited CT scan to rule out the possibility of a deeper abscess. One week after FL and wound assessment, the symptoms of LOWER infection had subsided and most UPPER symptoms dramatically improved, with pain being the only UPPER symptom that persisted. Fluorescence imaging at this time point indicated persistence of red/blush fluorescence but to a much lesser extent that previously observed. Fluorescence-guided debridement resulted in a reduction in the level of red fluorescence, indicative of bacteria, observed in the wound.