Amputation Wound Healing Research Articles

Machine learning analysis of multispectral imaging and clinical risk factors to predict amputation wound healing.

Prediction of amputation wound healing is challenging due to the multifactorial nature of critical limb ischemia and lack of objective assessment tools. Up to one-third of amputations require revision to a more proximal level within 1year. We tested a novel wound imaging system to predict amputation wound healing at initial evaluation. Patients planned to undergo amputation due to critical limb ischemia were prospectively enrolled. Clinicians evaluated the patients in traditional fashion, and all clinical decisions for amputation level were determined by the clinician's judgement. Multispectral images of the lower extremity were obtained preoperatively using a novel wound imaging system. Clinicians were blinded to the machine analysis. A standardized wound healing assessment was performed on postoperative day 30 by physical exam to determine whether the amputation site achieved complete healing. If operative revision or higher level of amputation was required, this was undertaken based solely upon the provider's clinical judgement. A machine learning algorithm combining the multispectral imaging data with patient clinical risk factors was trained and tested using cross-validation to measure the wound imaging system's accuracy of predicting amputation wound healing. A total of 22 patients undergoing 25 amputations (10 toe, five transmetatarsal, eight below-knee, and two above-knee amputations) were enrolled. Eleven amputations (44%) were non-healing after 30days. The machine learning algorithm had 91% sensitivity and 86% specificity for prediction of non-healing amputation sites (area under curve, 0.89). This pilot study suggests that a machine learning algorithm combining multispectral wound imaging with patient clinical risk factors may improve prediction of amputation wound healing and therefore decrease the need for reoperation and incidence of delayed healing. We propose that this, in turn, may offer significant cost savings to the patient and health system in addition to decreasing length of stay for patients.

Learning from machine learning.

In this article, Squiers et al1 bring together two different ideas to help address the salient clinical problem of successful amputation wound healing. In this pilot study looking at the use of multispectral imaging (MSI) in predicting wound healing, the authors have introduced machine learning (ML) to develop an ideal prediction algorithm. Hidden underneath these ideas are the clinical factors surgeons knowingly or unknowingly use to determine when and where to amputate. I suggest that the readers interpret these findings while keeping in mind that the ultimate goal of a successful amputation is not just healing, but maintaining functionality of the longest residual limb possible.

Machine Learning Analysis of Multispectral Imaging and Clinical Risk Factors to Predict Amputation Wound Healing

Machine Learning Analysis of Multispectral Imaging and Clinical Risk Factors to Predict Amputation Wound Healing

Do Pre-Operative Transcutaneous Oxygen Perfusion Measurements Predict Atraumatic Major Lower Extremity Amputation Wound Healing?

Category:OtherIntroduction/Purpose:Determining appropriate amputation level is a challenging task requiring clinical, laboratory, and imaging data. However, there is no validated consensus on the method with the most prognostic accuracy. Transcutaneous oxygen perfusion measurement (TcPO2) is a noninvasive means of measuring tissue oxygenation. A TcPO2 > 30-40mm Hg is widely cited as a positive predictor of post-operative wound healing, but its validity has not been well defined. We hypothesized that TcPO2 levels positively correlate with the success of amputation wound healing. Additionally, we secondarily investigated the correlation between other preoperative demographics and clinical variables and their impact on post-operative amputation wound healing.Methods:A retrospective chart review was performed on patients who underwent lower extremity above, through, or below knee amputations at a single institution with documented preoperative TcPO2 values between January 1, 2012 and December 1, 2018 and a minimum 30 days post-operative clinical follow up. Amputations performed for oncologic pathology, infected arthroplasty, osteomyelitis and traumatic amputations were excluded in order to isolate TcPO2 as a decision-making test for amputation level. This yielded one hundred and forty-one total amputations. Of these ninety-three were below knee amputations (BKA), six through knee amputations, and forty-two above knee amputations (AKA). Chi-square and t-tests were used to compare successful and failed amputations where appropriate.Results:Eighty-six of the amputations were successful and fifty-five failed, as defined by post-operative wound dehiscence or infection. Of these, thirty-seven were BKAs, four were through knee amputations and fourteen were AKAs. There was a significant difference in preoperative TcPO2 levels between the successful and failure groups at 46.2 and 38.3 respectively (p = 0.02). A TcPO2 of 30-40mm Hg had a success rate of 68.8%, and a TcPO2 < 20mm Hg a success rate of 18.2%. A receiver operating characteristic curve for TcPO2 levels predicting amputation success elucidated that with an area under the curve of 0.53 for the AKA cohort and 0.61 for the BKA cohort, the diagnostic ability is far from prognostic.Conclusion:Our results provide new insight into the predictive accuracy of preoperative TcPO2 levels. There is not a linear association between TcPO2 and success rate. A TcPO2 < 20mm Hg has a high positive predictive value for failure, but higher TcPO2 levels are not 100% predictive of amputation wound healing as reported by prior studies. Multiple factors should be considered when selecting amputation level.

Influencing Factors of Outcome After Lower-Limb Amputation: A Five-Year Review in a Plastic Surgical Department

The crude major lower limb amputation procedure rate is 8.8 per 100,000 of the population per year in Taiwan. From January 2002 to October 2006, patients that received major lower limb amputation in our department were enrolled in this study. Retrospective chart reviews concerning different factors that can affect the eventual postoperative functional status were investigated. Factors that affected the length of hospital stay included duration before amputation (P < 0.001) and renal function (P = 0.045). Phantom limb pain was affected by wound healing time (P = 0.006). Factors that affected the daily prosthesis usage time were initial infection status (P = 0.021), renal function (P = 0.01), patient educational level (P = 0.016), and pretraining waiting time (P = 0.003). The duration of prosthetic training was affected by patient educational level (P = 0.004) and marital status (P = 0.024). In addition, subjective satisfaction about the usage of prosthesis was affected by pretraining waiting time (P = 0.001) and daily prosthesis usage time (P < 0.001). The daily prosthesis usage time was closely related to life quality improvement (P < 0.001) and subjective satisfaction of prosthesis usage (P < 0.001). Despite reported unchangeable factors like age, end-stage renal disease, dementia, coronary artery disease, and level of amputation, preprosthesis training waiting time significantly affected the satisfaction and daily usage time of the prosthesis. Surgeons can make some contribution to accelerate amputation wound healing and stump maturation by choosing the correct operating procedure, delicately managing the soft tissue, and ascertaining proper wound care to improve the outcome.

Guidelines for the treatment of arterial insufficiency ulcers

1. Co-chaired panel 2. University of Utah, Salt Lake City, UT 3. University of Texas, San Antonio, TX 4. Sinai Hospital/Johns Hopkins Medical Institutions, Baltimore, MD 5. GKT School of Medicine, King’s College, London, UK 6. University of Texas Health Science Center at Houston, TX 7. Sequoia Hospital, Redwood City, CA 8. Maricopa Medical Center, Phoenix, AZ 9. Tufts-New England Medical Center, Boston, MA 10. Southampton University Hospitals Trust NHS, Southampton, UK 11. Penrose–St. Francis Health Services, Colorado Springs, CO 12. Beverly Surgical Associates, Beverly, MA 13. Saint Francis Memorial Hospital, San Francisco, CA 14. Northbay Center for Wound Care, Vacaville, CA, and 15. University of California, San Francisco, CA

Prediction of amputation wound healing: the role of transcutaneous pO2 assessment.

With the recent trend towards more distal lower limb and below-knee amputation for peripheral vascular disease, failure of amputation healing remains a common clinical problem. There is an urgent need for more objective measures of selecting the most appropriate distal amputation level compatible with healing. Oxygen availability is the final arbiter of tissue viability and healing potential, and we have shown that the measurement of transcutaneous pO2 (TcpO2) accurately reflects the degree of ischaemia in the lower limb. In 59 patients having 62 amputations for peripheral vascular disease significantly lower TcpO2 levels were related to amputation failure. Below-knee amputations with a pre-operative below-knee TcpO2 of above 35 mmHg always healed, and failures had levels of 35 mmHg or less. Nine out of sixteen patients having above-knee amputations had pre-operative below-knee TcpO2 values well above 35 mmHg, suggesting that they may possibly have undergone successful below-knee amputation. No correlation between ankle systolic pressure and amputation healing was found. TcpO2 measurement is simple, non-invasive and reliable and offers an exciting research advance in the assessment of effective tissue perfusion.

Factors influencing the healing of distal amputations performed for lower limb ischaemia

A total of 235 toes were amputated during 125 operations on 100 consecutive patients with lower limb ischaemia. The overall amputation wound healing rate for the series was 58.4 per cent and limb salvage was achieved in 66 patients. Toe amputation was performed under local anaesthesia in 57 cases and 32 (56 per cent) of these healed primarily, not significantly different from the healing rate of 41 (60 per cent) of 68 under general anaesthesia. Reconstructive arterial surgery was performed in conjunction with toe amputation in 39 patients; the healing rate with reconstruction was 32 (82 per cent) of 39, significantly better than the 41 (48 per cent) of 86 patients not undergoing bypass surgery (P less than 0.001). There was no difference in healing rates when comparing diabetic and non-diabetic patients. Multiple regression analysis demonstrated that reconstructive arterial surgery was the only factor which had an independent and significant influence on toe amputation healing. The use of local anaesthesia for distal amputation has no deleterious effects on wound healing.

Prediction of amputation wound healing with skin perfusion pressure

The purpose of this study was to determine whether laser Doppler skin perfusion pressure (LD-SPP) could accurately predict amputation wound healing. We studied a total of 62 limbs in 52 patients (28 men and 24 women) with a mean age of 62.2 years (range 34 to 93 years). From this, 39 limbs underwent major amputation (15 above-the-knee, 24 below-the-knee), and 23 limbs underwent minor amputations (4 transmetatarsal and 19 toes). There were five postoperative deaths, leaving a total of 57 limbs available for analysis. Three of 13 above-knee amputations failed to heal. Twenty-one of 23 below-knee amputations healed. Three of four transmetatarsal amputations failed to heal and eight of 17 toe amputations failed to heal. Binary table analysis showed that an LD-SPP value of 30 mm Hg or greater had a negative predictive value (healing occurred) of 90%. An LD-SPP value of less than 30 mm Hg at the amputation site had a positive predictive value (healing failure) of 75%, (p < 0.001, chi square analysis). For major amputations, negative predictive value was 100%, and positive predictive value was 83%, (p < 0.001). For minor amputations, negative predictive value was 75% and positive predictive value was 66.7%, (p < 0.09). These data support the use of the LD-SPP test in the selection of major amputation level consistent with healing in ischemic limbs. Further study of the value of this parameter in the determination of minor amputation wound healing is necessary.

Postoperative transcutaneous oxygen measurement in the prediction of delayed wound healing and prosthetic fitting among amputees during rehabilitation. A pilot study.

Postoperative assessment of amputation wound healing remains largely subjective in nature, being based on the physician's clinical judgement. These considerations significantly impact on the rehabilitation course, as premature prosthetic fitting may result in wound breakdown. Alternatively, delayed healing may result in prolonged hospital length of stay. Few attempts have been made to correlate objective parameters of limb perfusion with amputation wound healing or prosthetic fitting outcome during the rehabilitation phase of treatment. A pilot study was conducted, in which the transcutaneous oxygen monitor, a noninvasive device measuring transcutaneous partial pressure of oxygen (tcpO2), was applied to the stumps of 11 consecutive above-or below-knee amputees admitted for rehabilitation after amputation. All patients were tested within 1 wk of admission and 45 days of amputation. The treatment team was blinded as to the test results. A direct correlation was observed between wound healing outcome and tcpO2 results (Fisher's exact test [FET], P = 0.03), and no patient with a tcpO2 of < or 15 mm Hg healed during their rehabilitation stay (FET, P = 0.006). TcpO2 of < or = 15 mm Hg was significantly correlated with prolonged length of stay (Point Biserial Correlation Coefficient [rpbi], = -0.835; P = 0.01), delayed prosthetic fitting (rpbi = 0.742; p = 0.01), and poorer wound healing at admission (rpbi = 0.932; P = 0.001). Postoperative tcpO2 measurement may have use in objectively identifying patients at greater risk of delayed wound healing and prosthetic fitting, although further study is warranted.

Transcutaneous oxygen as a predictor of wound healing in amputations of the foot and ankle.

Thirty-eight amputations of the foot and ankle were performed in patients with peripheral vascular insufficiency over a 20-month period. Amputation level selection was based on clinical examination, a minimum ankle-brachial index of 0.5 as a measure of vascular supply, serum albumin of 3.0 gm/dl as a measure of tissue nutrition, and a total lymphocyte count of 1500 as a measure of immunocompetence. Transcutaneous oxygen tension was measured at the midfoot and ankle levels prior to surgery. Thirty-two of 38 patients (84.2%) healed their amputation wounds. When the transcutaneous oxygen tension was greater than 30 mm Hg, 24 of 26 patients (92.3%) healed. When the value was below 30 mm Hg, only eight of 12 patients healed. When the propensity to support wound healing is factored out, with patients having the metabolic capacity to heal an amputation wound in the foot and ankle, it appears that transcutaneous oxygen tension is an accurate measure of vascular inflow to support amputation wound healing.

Prediction of below-knee amputation wound healing using noninvasive laser Doppler velocimetry

Preoperative laser Doppler velocimetry (LDV) measurements of calf and brachial skin perfusion were performed in 29 patients prior to undergoing below-knee amputation; dual calf measurements on the anterior and posterior skin flaps were routinely obtained. Patients' ages ranged from 24 to 83 years; 16 had diabetes, 13 were smokers, 6 had hypertension, and 2 had known coronary artery disease. Anterior and posterior calf LDV values greater than or equal to 20 mV were associated with successful below-knee amputation wound healing in 25 of 26 patients; all 3 patients with either anterior or posterior calf LDV values less than 20 mV had below-knee amputations that failed to heal. Calculation of calf-brachial LDV indexes did not increase predictive accuracy compared to calf LDV measurements alone, although patients with wounds that failed to heal tended to have lower calf-brachial indexes. Preoperative noninvasive LDV may be used to aid in predicting the likelihood of successful healing after below-knee amputation.

Multisensor transcutaneous oximetric mapping to predict below-knee amputation wound healing: use of a critical Po2.

Preoperative transcutaneous oxygen tension (tcPo2) measurements of calf and brachial skin were performed on 40 patients before they underwent below-knee amputation (BKA); dual calf measurements on the anterior and posterior skin were obtained. Both anterior and posterior calf tcPo2 values were significantly lower in patients with unsuccessful wound healing after BKA compared to those with successful healing (p less than 0.01). Successful wound healing occurred in 50% (6/12) of patients with calf tcPo2 values less than 20 torr and in 96% (27/28) of patients with calf tcPo2 values greater than 20 torr. Calculation of a critical Po2 index (defined as the lesser of the anterior and posterior calf/brachial tcPo2 ratios) resulted in improved predictive accuracy; 100% (6/6) of patients with a critical Po2 index of 0.20 or less had unsuccessful wound healing after BKA, whereas 97% (33/34) of patients with a critical Po2 index greater than 0.20 had successful healing. The use of a critical Po2 index greater than 0.20 as predictive of successful healing after BKA was associated with a sensitivity, specificity, and overall accuracy of 100%, 86%, and 98%, respectively, compared to 82%, 86%, and 83%, respectively, by use of absolute calf tcPo2 values. In conclusion, multisensor transcutaneous oximetric mapping is an accurate method for predicting wound healing success after BKA. Measurement of both anterior and posterior calf tcPo2 should be performed; calf tcPo2 values less than 20 torr may indicate local ischemia, but ischemia should be confirmed by comparison of calf tcPo2 with brachial tcPo2 before the patient is denied BKA.(ABSTRACT TRUNCATED AT 250 WORDS)

Skin fluorescence, a method to predict amputation site healing

Although skin blood flow is an important determinant of amputation site healing, most techniques to evaluate it are cumbersome and indirect. We have prospectively compared a new, simple and direct technique, skin fluorescence (SF), with a more established indirect method, segmental limb blood pressures (BP), to determine the value of each in predicting amputation healing of an ischemic extremity. Twenty-six consecutive patients with ischemic tissue loss were studied prior to amputation. Thigh, calf, and ankle BP were measured; then SF was determined by injecting 10–15 mg/kg of fluorescein intravenously and examining the limb with uv light. BP of > 50 mm Hg was considered adequate for amputation wound healing; yellow SF indicated adequate skin blood flow. In each case, the most distal site predicted to heal by either method was selected for amputation. Thirty amputations were performed (6 digital, 3 transmetatarsal, 11 below knee, and 10 above knee) with an 80% healing rate. When SF and BP disagreed on the level of amputation, fluorescein always predicted a more distal site. SF accurately predicted amputation healing in 80% of cases. In the remaining 20%, amputation failures were due to infection in two and deep tissue necrosis in four. BP was accurate and useful in 47% of cases. In 14 patients BP indicated the need for a higher level of amputation than was successfully performed. In two patients BP measurement was indeterminate due to “stiff arteries.” Furthermore, in two patients amputations failed despite BP > 50 mm Hg. We conclude that SF helps the surgeon select a more distal level of amputation than segmental limb pressures, and one that has a high likelihood of healing.

Prediction of amputation wound healing. Roles of Doppler ultrasound and digit photoplethysmography.

Segmental limb blood pressures (BPs) measured by Doppler ultrasound and digit photoplethysmography were used to define hemodynamic correlates of wound healing in 122 lower-extremity amputations for ischemia. Healing of digit or transmetatarsal amputations were influenced by the presence of diabetes and correlated better with digit pulsation and BP than with ankle pressure. Healing of below-knee amputation did not correlate with segmental limb BPs. In the absence of an absolute threshold to predict wound healing or failure, caution is urged in using hemodynamic indices to select the level of amputation of arterial insufficiency.