This study was done to determine the incidence and patterns of split thickness skin graft failure at Kiruddu National Referral Hospital in Uganda. We enrolled 160 study participants, and none was lost during follow up. Majority were female with a mean age of 32.5 years and came from urban areas. The fact that majority came from urban areas was expected in this study, given that the study centre is located in the capital. Therefore most of the participants were more likely to come from within the city except for those referred with complex conditions from lower level facilities in the rural areas.
The commonest indications for grafting were traumatic ulcer and burn ulcer. These contributed the majority because the main causes of wounds that may require grafting are burns and trauma. Close to half of the participants had hypoalbuminemia. This is an alarming finding, however it can be explained by the fact that most of the patients in the government national referral are financially limited, since the rich usually go to private hospitals. The financial limitations are likely to cause inadequate intake, resulting in low albumin. More so, some of these patients were bed ridden for a long time resulting in some degree of malnutrition.
We noted that, 59 participants had graft failure representing an incidence ratio of 36.9% (369 graft failures per 1000 grafts done). In agreement with our findings, a study done in Pakistan reported that, of the 79 patients that underwent grafting, 35% experienced graft failure (9) and a study that enrolled 70 patients in New Zealand who underwent lower limb skin grafts reported that one-third of lower limb skin grafts failed (6). All these studies that had failure rates close the failure rate seen in this study included patients that had grafting of the lower limbs. Since the anatomical site grafted has a large bearing on the risk of failure, the similarity in our findings can be explained by the fact that all our studies involved grafting the lower limbs.
Contrally to our findings, among patients that had day case surgeries in Australia, the split-skin grafting had a failure rate of 66% when done in the lower limbs (14) The exact explanation for the higher rates seen in this Australia study is not known, however the fact that these patients were managed as day cases (discharged the same day of the surgery) could have played a role in the outcome, more so if the patient could have gone home without proper instructions or cared for postoperatively by a health work that is not well versed with the post-operative procedure for grafting.
The main direct cause of failure was infection. The commonest causative organism was Pseudomonas aeruginosa the others being Staphylococcus hemolyticus, Proteus mirabilis and Staphylococcus aureus. The sensitivities varied widely, however Amikacin had the highest sensitivity against all the organisms isolated in this study, but all organisms were resistant to ampicillin, cotrimoxazole and ceftriaxone.
Contrally to our findings where no patient had compete failure, a research that included diabetic patients in Arizona, USA reported that five (5.3%) patients ultimately required major limb amputation after complete graft failure (15). In India, out of 50 cases studied, graft rejection (complete failure) occurred in 6% (16). The reason for having no complete failure in our study could be because we assessed failure of the graft on the 10th post-operative day. If we had measured failure at a later day, possibly we could have had some patients with complete failure.
In agreement with our findings where infection was the main cause of failure, in Turkey, among one hundred thirty-two patients that underwent skin grafting, graft loss secondary to infection was recorded in 31 patients (23.5%) (13) and in Ghana, out of 57 infected wounds, 14 (24.6%) had infected grafted wounds that failed to heal. Infection is the commonest cause of failure possibly because other causes of failure can be easily predicted and grafting foregone or grating done and the complications prevented by appropriate dressing and splinting. However, colonisation and infection can occur at any time from the time of grafting to the post-operative period, making it less predictable.
Still in agreement, Pseudomonas aeruginosa was the commonest cause of infection as was reported by Kwesi et al (11) who noted that the bacteria that were frequently responsible for graft failure were identified as Pseudomonas aeruginosa and other Pseudomonas species. In this study we identified other species of bacteria among patients with failed grafts. The difference in the organisms isolated is because the pathological bacteria vary from hospital to hospital making it necessary for every hospital to establish an antibiogram to understand the commonest organisms and the antibiotics they respond to. We noted varying degrees of sensitivity, and this can be attributed to antibacterial stewardship that also varies from region to region.
Study limitation
We measured failure on the 10th postoperative day, and so failures that occurred later were not captured. Also, the source of infection was not identified as to whether it was from the operating room or the wards.