Prediction of Neurological Outcomes in Out-of-hospital Cardiac Arrest Survivors Immediately after Return of Spontaneous Circulation: Ensemble Technique with Four Machine Learning Models

Background@#We performed this study to establish a prediction model for 1-year neurological outcomes in out-of-hospital cardiac arrest (OHCA) patients who achieved return of spontaneous circulation (ROSC) immediately after ROSC using machine learning methods. @*Methods@#We performed a retrospective analysis of an OHCA survivor registry. Patients aged ≥ 18 years were included. Study participants who had registered between March 31, 2013 and December 31, 2018 were divided into a develop dataset (80% of total) and an internal validation dataset (20% of total), and those who had registered between January 1, 2019 and December 31, 2019 were assigned to an external validation dataset. Four machine learning methods, including random forest, support vector machine, ElasticNet and extreme gradient boost, were implemented to establish prediction models with the develop dataset, and the ensemble technique was used to build the final prediction model. The prediction performance of the model in the internal validation and the external validation dataset was described with accuracy, area under the receiver-operating characteristic curve, area under the precision-recall curve, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). Futhermore, we established multivariable logistic regression models with the develop set and compared prediction performance with the ensemble models. The primary outcome was an unfavorable 1-year neurological outcome. @*Results@#A total of 1,207 patients were included in the study. Among them, 631, 139, and 153were assigned to the develop, the internal validation and the external validation datasets, respectively. Prediction performance metrics for the ensemble prediction model in the internal validation dataset were as follows: accuracy, 0.9620 (95% confidence interval [CI],0.9352–0.9889); area under receiver-operator characteristics curve, 0.9800 (95% CI, 0.9612– 0.9988); area under precision-recall curve, 0.9950 (95% CI, 0.9860–1.0000); sensitivity, 0.9594 (95% CI, 0.9245–0.9943); specificity, 0.9714 (95% CI, 0.9162–1.0000); PPV, 0.9916 (95% CI, 0.9752–1.0000); NPV, 0.8718 (95% CI, 0.7669–0.9767). Prediction performance metrics for the model in the external validation dataset were as follows: accuracy, 0.8509 (95% CI, 0.7825–0.9192); area under receiver-operator characteristics curve, 0.9301 (95% CI, 0.8845–0.9756); area under precision-recall curve, 0.9476 (95% CI, 0.9087–0.9867); sensitivity, 0.9595 (95% CI, 0.9145–1.0000); specificity, 0.6500 (95% CI, 0.5022–0.7978); PPV, 0.8353 (95% CI, 0.7564–0.9142); NPV, 0.8966 (95% CI, 0.7857–1.0000). All the prediction metrics were higher in the ensemble models, except NPVs in both the internal and the external validation datasets. @*Conclusion@#We established an ensemble prediction model for prediction of unfavorable 1-year neurological outcomes in OHCA survivors using four machine learning methods. The prediction performance of the ensemble model was higher than the multivariable logistic regression model, while its performance was slightly decreased in the external validation dataset.

Prediction of Neurological Outcomes in Out-of-hospital Cardiac Arrest Survivors Immediately after Return of Spontaneous Circulation: Ensemble Technique with Four Machine Learning Models

Background@#We performed this study to establish a prediction model for 1-year neurological outcomes in out-of-hospital cardiac arrest (OHCA) patients who achieved return of spontaneous circulation (ROSC) immediately after ROSC using machine learning methods. @*Methods@#We performed a retrospective analysis of an OHCA survivor registry. Patients aged ≥ 18 years were included. Study participants who had registered between March 31, 2013 and December 31, 2018 were divided into a develop dataset (80% of total) and an internal validation dataset (20% of total), and those who had registered between January 1, 2019 and December 31, 2019 were assigned to an external validation dataset. Four machine learning methods, including random forest, support vector machine, ElasticNet and extreme gradient boost, were implemented to establish prediction models with the develop dataset, and the ensemble technique was used to build the final prediction model. The prediction performance of the model in the internal validation and the external validation dataset was described with accuracy, area under the receiver-operating characteristic curve, area under the precision-recall curve, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). Futhermore, we established multivariable logistic regression models with the develop set and compared prediction performance with the ensemble models. The primary outcome was an unfavorable 1-year neurological outcome. @*Results@#A total of 1,207 patients were included in the study. Among them, 631, 139, and 153were assigned to the develop, the internal validation and the external validation datasets, respectively. Prediction performance metrics for the ensemble prediction model in the internal validation dataset were as follows: accuracy, 0.9620 (95% confidence interval [CI],0.9352–0.9889); area under receiver-operator characteristics curve, 0.9800 (95% CI, 0.9612– 0.9988); area under precision-recall curve, 0.9950 (95% CI, 0.9860–1.0000); sensitivity, 0.9594 (95% CI, 0.9245–0.9943); specificity, 0.9714 (95% CI, 0.9162–1.0000); PPV, 0.9916 (95% CI, 0.9752–1.0000); NPV, 0.8718 (95% CI, 0.7669–0.9767). Prediction performance metrics for the model in the external validation dataset were as follows: accuracy, 0.8509 (95% CI, 0.7825–0.9192); area under receiver-operator characteristics curve, 0.9301 (95% CI, 0.8845–0.9756); area under precision-recall curve, 0.9476 (95% CI, 0.9087–0.9867); sensitivity, 0.9595 (95% CI, 0.9145–1.0000); specificity, 0.6500 (95% CI, 0.5022–0.7978); PPV, 0.8353 (95% CI, 0.7564–0.9142); NPV, 0.8966 (95% CI, 0.7857–1.0000). All the prediction metrics were higher in the ensemble models, except NPVs in both the internal and the external validation datasets. @*Conclusion@#We established an ensemble prediction model for prediction of unfavorable 1-year neurological outcomes in OHCA survivors using four machine learning methods. The prediction performance of the ensemble model was higher than the multivariable logistic regression model, while its performance was slightly decreased in the external validation dataset.

Copeptin with high-sensitivity troponin at presentation is not inferior to serial troponin measurements for ruling out acute myocardial infarction

Objective@#We aimed to compare the multi-marker strategy (copeptin and high-sensitivity cardiac troponin I [hs-cTnI]) with serial hs-cTnI measurements to rule out acute myocardial infarction (AMI) in patients with chest pain. @*Methods@#This prospective observational study was performed in a single emergency department. To test the non-inferiority margin of 4% in terms of negative predictive value (NPV) between the multi-marker strategy (0 hour) and serial hs-cTnI measurements (0 and 2 hours), 262 participants were required. Samples for copeptin and hs-cTnI assays were collected at presentation (0 hour) and after 2 hours. The measured biomarkers were considered abnormal when hs-cTnI was >26.2 ng/L and when copeptin was >10 pmol/L. @*Results@#AMI was diagnosed in 28 patients (10.7%). The NPV of the multi-marker strategy was 100% (160/160; 95% confidence interval [CI], 97.7% to 100%), which was not inferior to that of serial hs-cTnI measurements (201/201; 100%; 95% CI, 98.2% to 100%). The sensitivity, specificity, and positive predictive value of the multi-marker strategy were 100% (95% CI, 87.7% to 100%), 68.1% (95% CI, 61.7% to 74.0%), and 27.2% (95% CI, 18.9% to 36.8%), respectively. The sensitivity, specificity, and positive predictive value of serial hs-cTnI measurements were 100% (95% CI, 87.7% to 100%), 85.5% (95% CI, 80.4% to 89.8%), and 45.2% (95% CI, 32.5% to 58.3%), respectively. @*Conclusion@#The multi-marker strategy (copeptin and hs-cTnI measurement) was not inferior to serial hs-cTnI measurements in terms of NPV for AMI diagnosis, with a sensitivity and NPV of 100%. Copeptin may help in the early rule-out of AMI in patients with chest pain.

Treatment Outcomes and Response Pattern of Ustekinumab in Korean Patients with Psoriasis: A Retrospective Single-center Study / 대한피부과학회지

BACKGROUND: Psoriasis is a chronic immune-mediated inflammatory skin disease affecting 2~3% of the worldwide population. Ustekinumab, an IL-12/23p40 inhibitor, is a biologic reported to be effective and safe in treating psoriasis. However, there are limited data on the treatment outcomes of ustekinumab in patients with psoriasis in Korea. OBJECTIVE: To evaluate the treatment outcomes and response pattern of ustekinumab in patients with psoriasis in Korea. METHODS: This was a retrospective single-center study. Eighty-four patients with psoriasis treated with ustekinumab were analyzed. Each patient's medical records, psoriasis area and severity index (PASI) score, and body surface area were reviewed at baseline and up to week 52. RESULTS: A total of 84 patients were included (male:female=1.8:1). The mean age was 44.5 years. At week 16, 86.7% achieved PASI75, 59.0% achieved PASI90, and 20.5% achieved PASI100. By week 16, 84.8% of subjects had attained PASI75 for the head region, whereas 79.0% had attained it for the lower extremities, indicating a relatively slower treatment response of psoriatic lesions on the lower extremities. Four patients discontinued treatment due to lack of effect. No severe adverse events occurred during the follow-up period. CONCLUSION: Ustekinumab demonstrated highly effective and safe treatment profiles in Korean psoriatic patients, consistent with the previous reports from mainly Western countries. Psoriasis severity and treatment responsiveness may vary with body region.

Association between the simultaneous decrease in the levels of soluble vascular cell adhesion molecule-1 and S100 protein and good neurological outcomes in cardiac arrest survivors

OBJECTIVE: This study aimed to determine whether simultaneous decreases in the serum levels of cell adhesion molecules (intracellular cell adhesion molecule-1 [ICAM-1], vascular cell adhesion molecule-1 [VCAM-1], and E-selectin) and S100 proteins within the first 24 hours after the return of spontaneous circulation were associated with good neurological outcomes in cardiac arrest survivors. METHODS: This retrospective observational study was based on prospectively collected data from a single emergency intensive care unit (ICU). Twenty-nine out-of-hospital cardiac arrest survivors who were admitted to the ICU for post-resuscitation care were enrolled. Blood samples were collected at 0 and 24 hours after ICU admission. According to the 6-month cerebral performance category (CPC) scale, the patients were divided into good (CPC 1 and 2, n=12) and poor (CPC 3 to 5, n=17) outcome groups. RESULTS: No difference was observed between the two groups in terms of the serum levels of ICAM-1, VCAM-1, E-selectin, and S100 at 0 and 24 hours. A simultaneous decrease in the serum levels of VCAM-1 and S100 as well as E-selectin and S100 was associated with good neurological outcomes. When other variables were adjusted, a simultaneous decrease in the serum levels of VCAM-1 and S100 was independently associated with good neurological outcomes (odds ratio, 9.285; 95% confidence interval, 1.073 to 80.318; P=0.043). CONCLUSION: A simultaneous decrease in the serum levels of soluble VCAM-1 and S100 within the first 24 hours after the return of spontaneous circulation was associated with a good neurological outcome in out-of-hospital cardiac arrest survivors.

Admission levels of high-density lipoprotein and apolipoprotein A-1 are associated with the neurologic outcome in patients with out-of-hospital cardiac arrest

OBJECTIVE: To investigate whether serum levels of high-density lipoprotein (HDL) and apolipoprotein A-1 (ApoA1), after the return of spontaneous circulation, can predict the neurologic outcome in patients with out-of-hospital cardiac arrest (OHCA). METHODS: This was a retrospective observational study conducted in a single tertiary hospital intensive care unit. All adult OHCA survivors with admission lipid profiles were enrolled from March 2013 to December 2015. Good neurologic outcome was defined as discharge cerebral performance categories 1 and 2. RESULTS: Among 59 patients enrolled, 13 (22.0%) had a good neurologic outcome. Serum levels of HDL (56.7 vs. 40 mg/dL) and ApoA1 (117 vs. 91.6 mg/dL) were significantly higher in patients with a good outcome. Areas under the HDL and ApoA1 receiver operating curves to predict good outcomes were 0.799 and 0.759, respectively. The proportion of good outcome was significantly higher in patients in higher tertiles of HDL and ApoA1 (test for trend, both P=0.003). HDL (P=0.018) was an independent predictor in the multivariate logistic regression model. CONCLUSION: Admission levels of HDL and ApoA1 are associated with neurologic outcome in patients with OHCA. Prognostic and potential therapeutic values of HDL and ApoA1 merit further evaluation in the post-cardiac arrest state, as in other systemic inflammatory conditions such as sepsis.

Anti-Cyclic Citrullinated Peptide Antibodies and Joint Involvement in Behcet's Disease

PURPOSE: We aimed to determine the prevalence of anti-cyclic citrullinated peptide (anti-CCP) antibodies in a large group of Korean patients with Behcet's disease (BD), with and without joint involvement, and to compare these findings with the prevalences of anti-CCP antibodies in patients with rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE). MATERIALS AND METHODS: We tested 189 patients with BD, 105 with RA, and 36 with SLE for anti-CCP antibodies and IgM rheumatoid factor in serum. We reviewed the medical records of patients with BD to investigate their personal and clinical characteristics as well as their laboratory test results. RESULTS: Anti-CCP antibodies were detected in seven of the 189 BD patients (3.7%), at a mean titer of 30.6+/-44.4 U/mL, in 86 of the 105 RA patients (81.9%) with a mean titer of 198.8+/-205.7 U/mL, and in nine of the 36 SLE patients (25%) with a mean titer of 180.4+/-113.9 U/mL. One of the seven anti-CCP-positive BD patients fulfilled the diagnostic criteria for both BD and RA. Five of the seven anti-CCP-positive BD patients (71.4%) had polyarticular joint involvement, and the other two patients (28.6%) had oligoarticular involvement. CONCLUSION: We determined the prevalence of anti-CCP antibodies in a large group of Korean BD patients with and without joint involvement. Negative anti-CCP test in patients with BD may help to differentiate BD from RA and SLE, all of which present with similar clinical features.

Effect of Emergency Department Video-nystagmography for Patients with Dizziness

PURPOSE: We evaluated the effects of emergency department (ED)-based video-nystagmography (VNG) testing on consultation with other departments and length of ED stay of patients with dizziness. METHODS: A before-and-after study was performed at a tertiary ED from May 13, 2011 to May 12, 2012. Adult patients (age> or =15 years) with dizziness were enrolled, excluding patients with incomplete information. We defined the before- and after-phase according to implanting of the ED-based VNG test. The VNG test was performed by an internship physician supervised by 2nd or 3rd grade emergency medicine residents. Primary outcome was any consultation to other specialty departments and the secondary outcome was the length of stay at the ED. The adjusted odds ratios (ORs) with 95% confidence interval (95% CI) for outcomes by phase were estimated using multivariate logistic regression analysis adjusting for potential co-variates. RESULTS: Of 1,485 eligible patients, 1,449 patients (male: 37.0%, mean age: 59.2+/-15.5 years) were enrolled (415 in the before-phase and 1,034 in the after-phase). The final diagnosis group was stroke (4.8%), peripheral vestibulopathy (44.9%), other specific disease (29.0%), and non-diagnostic symptom (21.4%). The consultation request was more reduced in the after-phase (38.1%) than the before-phase (52.5%). The LOS was not changed between both phases (7.4+/-7.5 hours versus 7.4+/-7.3 hours, p=0.76). The adjusted OR (95% CI) for the consultation of after-phase compared to before-phase was 0.46(0.35, 0.61), while the adjusted OR (95% CI) for LOS was 1.39(0.99, 1.95). CONCLUSION: Implementation of ED-based VNG test significantly reduced the consultation with other specialty departments, while LOS was not affected.

A Case of Subungual Epidermal Cyst / 대한피부과학회지

We report a 45-year-old woman presenting with a painful subungual epidermal cyst on the 4th finger nail bed. Based on clinical features only, the lesion was initially suspected to be a glomus tumor. However, histopathological examination revealed an epidermal cyst on the nail bed. Subungual epidermal cysts are rare disorders that may cause pain and tenderness that increase according to tumor growth. They can mimic glomus tumors owing to their similar clinical manifestations. To differentiate subungual epidermal cysts from glomus tumors, imaging tools such as ultrasonography and magnetic resonance imaging can be used. Dermatologists should be aware that several kinds of tumors including epidermal cysts with clinical similarities may occur in the subungual area.

A Case of Green Nail Syndrome in a Dress Designer with a Chronic Use of Thimble / 대한피부과학회지

Green nail syndrome is a chronic nail infection by Pseudomonas aeurginosa. Infected lesions can involve both finger and toe nails and cause greenish discoloration and onycholysis. Environmental factors such as frequent hydration and external trauma are predisposing factors. A 37-year-old patient was diagnosed with green nail syndrome on the right thumb. She used a thimble for a long time as a dress designer. Herein, we describe an uncommon case of green nail syndrome by chronic thimble use as a rare and an educational report.

Kasabach-Merritt Syndrome Arising from Tufted Angioma Successfully Treated with Systemic Corticosteroid

We report a case of Kasabach-Merritt syndrome arising from a tufted angioma successfully treated with systemic corticosteroid. A 2-month-old male infant presented with a palm-sized, erythematous induration on his left pubis. The lesion was diagnosed as tufted angioma histopathologically. After 1 month, the lesion suddenly expanded to the abdomen and scrotum. Initial laboratory tests were consistent with consumptive coagulopathy. He was diagnosed with Kasabach-Merritt syndrome and treated with intravenous dexamethasone at 0.32 mg/kg/day (equivalent to prednisolone 2.0 mg/kg/day). Two days after initiating the treatment, his platelet counts recovered and the lesion ceased to expand. Steroid therapy was converted to oral prednisolone and the dosage was subsequently tapered, and the lesion gradually involuted with no signs of recurrence for a year.

A comparative Study of SPECT with MRI in Internal Derangement of the Knee

Bone single photon emission computed tomography(SPECT) of the knees has been shown to be of diagnostic value for the investigation of intemal derangement of knees(IDK). This study compared bone SPECT and magnetic resonance image(MRI) with arthroscopic findings for the diagnosis of IDK. Fifty patients who had been performed arthroscopy in Seoul National University Hospital from January 1997 to September 1997 were studied with both the modalities preoperatively. There were 38 meniscal injuries, 26 cruciate ligament injuries and 8 cases of patellofemoral joint problems. In view of SPECT, the most remarkable results were from meniscal tears. In meniscal tears, the diagnostic values of SPECT were 89.5% of sensitivity, 66.7% of specificity, 84% of diagnostic accuracy, which were comparable to those of MRI, 97.4%, 91.7% and 96%, respectively. But SPECT showed inferior diagnostic values in cruciate ligament injuries. For anterior knee pain, SPECT was more sensitive than MRI for the detection of pathology. And for functionally related cases, such as ACL impingement to intercondylar notch, SPECT was far more valuable than MRI. Bone SPECT will not provide the anatomical detail of MRI but seems to be a useful tool for the detection of wide range of knee disorders including IDK. Further research is justified to investigate the precise role of bone SPECT in clinical practice and its value in relation to MRI.

Osteochondritis Dissecans of the Knee Associated with Gout: A Case Report

Osteochondritis dissecans of the knee occurs twice as often in males as in females. The peak incidence is about fifteen years of age, and most patients are first diagnosed before the age of eighteen. The etiology of osteochondritis dissecans is still unknown, but the most accepted theories are trauma, ischemia, abnormal ossification within the physes, constitutional or genetic predisposition, and a combination of these. Osteochondritis dissecans of the knee associated with gout was first reported by Carrabba in 1969. And no additional case has been reported. The purpose of this paper is to present an unusual case of osteochondritis dissecans of the knee associated with gout.