Emotion recognition in doctor-patient interactions from real-world clinical video database: Initial development of artificial empathy.

BACKGROUND AND OBJECTIVE: The promising use of artificial intelligence (AI) to emulate human empathy may help a physician engage with a more empathic doctor-patient relationship. This study demonstrates the application of artificial empathy based on facial emotion recognition to evaluate doctor-patient relationships in clinical practice. METHODS: A prospective study used recorded video data of doctor-patient clinical encounters in dermatology outpatient clinics, Taipei Municipal Wanfang Hospital, and Taipei Medical University Hospital collected from March to December 2019. Two cameras recorded the facial expressions of four doctors and 348 adult patients during regular clinical practice. Facial emotion recognition was used to analyze the basic emotions of doctors and patients with a temporal resolution of 1 second. In addition, a physician-patient satisfaction questionnaire was administered after each clinical session, and two standard patients gave impartial feedback to avoid bias. RESULTS: Data from 326 clinical session videos showed that (1) Doctors expressed more emotions than patients (t [326] > = 2.998, p < = 0.003), including anger, happiness, disgust, and sadness; the only emotion that patients showed more than doctors was surprise (t [326] = -4.428, p < .001) (p < .001). (2) Patients felt happier during the latter half of the session (t [326] = -2.860, p = .005), indicating a good doctor-patient relationship. CONCLUSIONS: Artificial empathy can offer objective observations on how doctors' and patients' emotions change. With the ability to detect emotions in 3/4 view and profile images, artificial empathy could be an accessible evaluation tool to study doctor-patient relationships in practical clinical settings.

Ten-year trends in depression care in Taiwan.

BACKGROUND/PURPOSE: The number of psychiatrists working in community clinics in Taiwan has increased dramatically in the recent decade. This study aimed to investigate the trend of prevalence and incidence of depressive disorders and assess the quality of depression care between 2007 and 2016 in Taiwan. METHODS: We used the claims database derived from Taiwan's National Health Insurance (NHI) program, in which approximately 23.0 million individuals were enrolled, translating to a coverage rate of 99%. Patients with depressive disorders were identified based on International Classification of Diseases codes. The process indicators of depression care quality included visit, duration, and dose adequacy. The outcome indicators included the rate of psychiatric hospitalisation, emergency visit, self-harm hospitalisation, and suicide. RESULTS: The prevalence of treated depressive disorders increased from 1.61% in 2007 to 1.92% in 2016, i.e., a 25% increase, whereas the incidence of first-ever or recurrent depressive disorder did not change significantly. The number of patients treated by psychiatrists and in community clinics also increased. The quality of depression care improved, the proportion of patients receiving minimum psychiatric clinic follow-up and adequate medication increased, and the rate of emergency visits, psychiatric hospitalisation, and self-harm hospitalisation declined. CONCLUSION: The community-based psychiatric services increased and the quality indicators of depression care in Taiwan improved during 2007-2016. The causality warrants further investigations.

Risk factors for poorly controlled and recurrence-prone ingrown toenails treated with nail braces: A retrospective observational study of 238 cases.

Background Nail braces are reportedly effective for treating both acute inflamed and chronic dystrophic type ingrown toenails. Aims In this study, risk factors for poorly controlled and recurrence-prone ingrown toenails treated with nail braces were identified. Methods We performed a retrospective study on patients with ingrown toenails between June 1, 2015, and May 31, 2018. The last follow-up date was January 31, 2019. Multivariate logistic regression was performed to evaluate the possible factors associated with poorly controlled status (ongoing paronychia during treatment) and recurrence. Results There were 120 (244 sides) and 118 patients (167 sides) with chronic dystrophic and acute inflamed type ingrown toenails, respectively. The mean treatment duration and follow-up period were 161.2 ± 98.3 days and 432.7 ± 320.9 days, respectively. Poor control and recurrence were seen in 7.3% (17/232) and 12.2% (27/221) of the patients, respectively. In the multivariate analysis, acute inflamed ingrown toenails, previous nail avulsion, proximal nail fold hypertrophy and more than one affected side remained significantly associated with poorly controlled ingrown toenails. Foot bone deformity was significantly associated with recurrence. Limitations This study was a retrospective study so that confounding factors such as comorbidities, body mass index, accompanying nail changes and lifestyle could not be evaluated. Conclusion Several risk factors related to poor control and recurrence were identified. Patients could therefore benefit from more suitable treatment plans with reasonable expectation.

Assessing Physicians' Recall Bias of Work Hours With a Mobile App: Interview and App-Recorded Data Comparison.

BACKGROUND: Previous studies have shown inconsistencies in the accuracy of self-reported work hours. However, accurate documentation of work hours is fundamental for the formation of labor policies. Strict work-hour policies decrease medical errors, improve patient safety, and promote physicians' well-being. OBJECTIVE: The aim of this study was to estimate physicians' recall bias of work hours with a mobile app, and to examine the association between the recall bias and physicians' work hours. METHODS: We quantified recall bias by calculating the differences between the app-recorded and self-reported work hours of the previous week and the penultimate week. We recruited 18 physicians to install the "Staff Hours" app, which automatically recorded GPS-defined work hours for 2 months, contributing 1068 person-days. We examined the association between work hours and two recall bias indicators: (1) the difference between self-reported and app-recorded work hours and (2) the percentage of days for which work hours were not precisely recalled during interviews. RESULTS: App-recorded work hours highly correlated with self-reported counterparts (r=0.86-0.88, P<.001). Self-reported work hours were consistently significantly lower than app-recorded hours by -8.97 (SD 8.60) hours and -6.48 (SD 8.29) hours for the previous week and the penultimate week, respectively (both P<.001). The difference for the previous week was significantly correlated with work hours in the previous week (r=-0.410, P=.01), whereas the correlation of the difference with the hours in the penultimate week was not significant (r=-0.119, P=.48). The percentage of hours not recalled (38.6%) was significantly higher for the penultimate week (38.6%) than for the first week (16.0%), and the former was significantly correlated with work hours of the penultimate week (r=0.489, P=.002). CONCLUSIONS: Our study identified the existence of recall bias of work hours, the extent to which the recall was biased, and the influence of work hours on recall bias.

Correction: Artificial Intelligence-Based Prediction of Lung Cancer Risk Using Nonimaging Electronic Medical Records: Deep Learning Approach.

[This corrects the article DOI: 10.2196/26256.].

Artificial Intelligence-Based Prediction of Lung Cancer Risk Using Nonimaging Electronic Medical Records: Deep Learning Approach.

BACKGROUND: Artificial intelligence approaches can integrate complex features and can be used to predict a patient's risk of developing lung cancer, thereby decreasing the need for unnecessary and expensive diagnostic interventions. OBJECTIVE: The aim of this study was to use electronic medical records to prescreen patients who are at risk of developing lung cancer. METHODS: We randomly selected 2 million participants from the Taiwan National Health Insurance Research Database who received care between 1999 and 2013. We built a predictive lung cancer screening model with neural networks that were trained and validated using pre-2012 data, and we tested the model prospectively on post-2012 data. An age- and gender-matched subgroup that was 10 times larger than the original lung cancer group was used to assess the predictive power of the electronic medical record. Discrimination (area under the receiver operating characteristic curve [AUC]) and calibration analyses were performed. RESULTS: The analysis included 11,617 patients with lung cancer and 1,423,154 control patients. The model achieved AUCs of 0.90 for the overall population and 0.87 in patients ≥55 years of age. The AUC in the matched subgroup was 0.82. The positive predictive value was highest (14.3%) among people aged ≥55 years with a pre-existing history of lung disease. CONCLUSIONS: Our model achieved excellent performance in predicting lung cancer within 1 year and has potential to be deployed for digital patient screening. Convolution neural networks facilitate the effective use of EMRs to identify individuals at high risk for developing lung cancer.

Serum Homocysteine, Folate, and Vitamin B12 Levels in Patients with Systemic Lupus Erythematosus: A Meta-Analysis and Meta-Regression.

BACKGROUND: Patients with systemic lupus erythematosus (SLE) have elevated cardiovascular risk. Hyperhomocysteinemia may be one of the contributing factors to this phenomenon. This study therefore aimed to compare the serum homocysteine levels and the levels of folate and vitamin B12, cofactors for homocysteine metabolism, between individuals with and without SLE. METHODS: A literature search was performed in PubMed, Embase, and the Cochrane library (from inception to March 31, 2019). Studies comparing serum homocysteine, folate or vitamin B12 levels between individuals with and without SLE were selected. Of the 1040 screened studies, 50 studies met the inclusion criteria. RESULTS: A total of 50 studies involving 4396 patients with SLE were included. Patients with SLE had a significantly higher serum level of homocysteine (standardized mean difference [SMD], 1.134; 95% CI, 0.795-1.474) and lower level of vitamin B12 (SMD, -0.359; 95% CI, -0.638 to -0.080) than controls. The folate level didn't differ markedly between SLE patients and the control group (SMD, -0.276; 95% CI, -0.674-0.123). Subgroup analysis showed consistent results in adult SLE patients. A random effects meta-regression analysis revealed a significantly inverse correlation between the SMD of homocysteine levels and C3 levels (coefficient, -0.0356, 95% CI, -0.054 to -0.0171; P < .001) and C4 levels (coefficient, -0.0876, 95% CI, -0.1407 to -0.0345; P = .0012). CONCLUSIONS: Serum homocysteine levels were higher and vitamin B12 levels were lower among individuals with SLE than those without SLE. Physicians are encouraged to monitor these parameters and offer timely interventions for patients with SLE.

Treating toxic epidermal necrolysis with systemic immunomodulating therapies: A systematic review and network meta-analysis.

BACKGROUND: Various systemic immunomodulating therapies have been used to treat toxic epidermal necrolysis (TEN), but their efficacy remains unclear. OBJECTIVE: To perform a systematic review and network meta-analysis (NMA) evaluating the effects of systemic immunomodulating therapies on mortality for Stevens-Johnson syndrome (SJS)/TEN overlap and TEN. METHODS: A literature search was performed in online databases (from inception to October 31, 2019). Outcomes were mortality rates and Score of Toxic Epidermal Necrolysis (SCORTEN)-based standardized mortality ratio (SMR). A frequentist random-effects model was adopted. RESULTS: Sixty-seven studies involving 2079 patients were included. An NMA of 10 treatments showed that none was superior to supportive care in reducing mortality rates and that thalidomide was associated with a significantly higher mortality rate (odds ratio, 11.67; 95% confidence interval [CI], 1.42-95.96). For SMR, an NMA of 11 treatment arms showed that corticosteroids and intravenous immunoglobulin combination therapy was the only treatment with significant survival benefits (SMR, 0.53; 95% CI, 0.31-0.93). LIMITATIONS: Heterogeneity and a paucity of eligible randomized controlled trials. CONCLUSIONS: Combination therapy with corticosteroids and IVIg may reduce mortality risks in patients with SJS/TEN overlap and TEN. Cyclosporine and etanercept are promising therapies, but more studies are required to provide clearer evidence.

Efficacy of Nail Braces for Acute and Chronic Ingrown Toenails: A Prospective Study.

BACKGROUND: Nail braces are an alternative treatment for ingrown toenails. OBJECTIVE: This study aimed to prospectively examine the efficacy of nail braces for treatment of acute inflamed (AI)-type and chronic dystrophic-type ingrown toenails. MATERIALS AND METHODS: The authors conducted a prospective study of patients with ingrown toenails treated at Wan Fang Hospital between January 1, 2017, and July 31, 2018. Evaluation using physician global assessment scores and patient satisfaction questionnaires was performed at 1, 3, and 6 months after the start of treatment and during the final visit. Patient demographics, treatment courses, and outcomes were compared between the 2 types of ingrown toenails. RESULTS: Chronic dystrophic-type and AI-type ingrown toenails were observed in 25 (61 sides) and 28 patients (35 sides), respectively. Of the affected sides, 80.9%, 94.9%, and 100% achieved an excellent or fair result at 1, 3, and 6 months, respectively. Treatment duration and follow-up period were 179.2 ± 96.8 days and 281.6 ± 120.9, respectively. The recurrence rate was 7.4%. The treatment course and response were different between the 2 types of ingrown toenails. CONCLUSION: Ingrown toenails could be effectively treated with nail braces with excellent outcomes, favorable patient satisfaction, and low recurrence rates.

Assessment of Deep Learning Using Nonimaging Information and Sequential Medical Records to Develop a Prediction Model for Nonmelanoma Skin Cancer.

IMPORTANCE: A prediction model for new-onset nonmelanoma skin cancer could enhance prevention measures, but few patient data-driven tools exist for more accurate prediction. OBJECTIVE: To use machine learning to develop a prediction model for incident nonmelanoma skin cancer based on large-scale, multidimensional, nonimaging medical information. DESIGN, SETTING, AND PARTICIPANTS: This study used a database comprising 2 million randomly sampled patients from the Taiwan National Health Insurance Research Database from January 1, 1999, to December 31, 2013. A total of 1829 patients with nonmelanoma skin cancer as their first diagnosed cancer and 7665 random controls without cancer were included in the analysis. A convolutional neural network, a deep learning approach, was used to develop a risk prediction model. This risk prediction model used 3-year clinical diagnostic information, medical records, and temporal-sequential information to predict the skin cancer risk of a given patient within the next year. Stepwise feature selection was also performed to investigate important and determining factors of the model. Statistical analysis was performed from November 1, 2016, to October 31, 2018. MAIN OUTCOMES AND MEASURES: Sensitivity, specificity, and area under the receiver operating characteristic (AUROC) curve were used to evaluate the performance of the models. RESULTS: A total of 1829 patients (923 women [50.5%] and 906 men [49.5%]; mean [SD] age, 65.3 [15.7] years) with nonmelanoma skin cancer and 7665 random controls without cancer (3951 women [51.5%] and 3714 men [48.4%]; mean [SD] age, 47.5 [17.3] years) were included in the analysis. The 1-year incident nonmelanoma skin cancer risk prediction model using sequential diagnostic information and drug prescription information as a time-incorporated feature matrix could attain an AUROC of 0.89 (95% CI, 0.87-0.91), with a mean (SD) sensitivity of 83.1% (3.5%) and mean (SD) specificity of 82.3% (4.1%). Carcinoma in situ of skin (AUROC, 0.867; -2.80% loss) and other chronic comorbidities (eg, degenerative osteopathy [AUROC, 0.872; -2.32% loss], hypertension [AUROC, 0.879; -1.53% loss], and chronic kidney insufficiency [AUROC, 0.879; -1.52% loss]) served as more discriminative factors for the prediction. Medications such as trazodone, acarbose, systemic antifungal agents, statins, nonsteroidal anti-inflammatory drugs, and thiazide diuretics were the top-ranking discriminative features in the model; each led to more than a 1% decrease of the AUROC when eliminated individually (eg, trazodone AUROC, 0.868; -2.67% reduction; acarbose AUROC, 0.870; -2.50 reduction; and systemic antifungal agents AUROC, 0.875; -1.99 reduction). CONCLUSIONS AND RELEVANCE: The findings of this study suggest that a risk prediction model may have potential predictive factors for nonmelanoma skin cancer. This model may help health care professionals target high-risk populations for more intensive skin cancer preventive methods.

A Novel, Optimized Method to Accelerate the Preparation of Injectable Poly-L-Lactic Acid by Sonication.

Current consensus for preparing injectable poly-L-lactic acid (PLLA) suggests adequate hydration (less than equal to 2-24 hours of reconstitution) of the lyophilized particles before injection, but the volume of reconstitution and the duration of hydration time varies. This study established a method to evaluate the distribution of PLLA particles after hydration and found that longer hydration time increased the effective portion (particles less than 60 µm) of PLLA products. Further investigation of the feasibility of reconstitution with sonication revealed that 2-hour hydration of PLLA powders with additional 5-minute-sonication could yield a comparable particle distribution with 48-hour-hydration of PLLA. Moreover, adding lidocaine into the diluent did not alter the distribution of PLLA particles. We proposed a new, feasible and efficient method of preparing PLLA injectable products: 2-hour hydration of the powders, sonication of the bottle or vial containing PLLA products for at least 5 minutes, and finalization with 1-2 mL of lidocaine immediately before injection. J Drugs Dermatol. 2018;17(8):894-898.

Whether CD44 is an applicable marker for glioma stem cells.

Glioblastoma multiforme (GBM) is one of the most malignant and aggressive brain tumors with great amount of hyaluronan (HA) secretion and CD44 overexpression (HA receptor). CD44 has been suggested as a cancer stem cells (CSCs) marker. However, several clinical studies have indicated that CD44low glioma cell exhibit CSCs traits. Additionally, our previous study indicated that more CD44 expression was associated with a better prognosis in GBM patients. To determine whether CD44 is an appropriate marker of glioma stem cells (GSCs), we manipulated CD44 expression using intrinsic (CD44 knockdown, CD44kd) and extrinsic (HA supplement, HA+) methods. Our results show that CD44kd suppressed cell proliferation by retarding cell cycle progression from G0/G1 to S phase. Furthermore, it caused GSCs traits, including lower expression of differentiation marker (glial fibrillary acidic protein, GFAP), a higher level of sphere formation and higher expression of stem cell markers (CD133, nestin and Oct4). The reduction of CD44 expression induced by HA+ was accompanied by an increase in GSCs properties. Interestingly, the presence of HA+ in glioma cells with GSC traits conversely facilitated differentiation. Our data indicated that the CD44 low-expressing cells exhibit more GSCs straits, suggesting that CD44 is not an appropriate marker for GSCs. Furthermore, the preferential expression of CD44 at the invasive rim in rat glioma specimen implies that CD44 may be more important for invasion and migration instead of GSCs marker in glioma.

GADD45A plays a protective role against temozolomide treatment in glioblastoma cells.

Glioblastoma multiforme (GBM) is one of the most aggressive cancers. Despite recent advances in multimodal therapies, high-grade glioma remains fatal. Temozolomide (TMZ) is an alkylating agent used worldwide for the clinical treatment of GBM; however, the innate and acquired resistance of GBM limits its application. Here, we found that TMZ inhibited the proliferation and induced the G2/M arrest of GBM cells. Therefore, we performed microarrays to identify the cell cycle- and apoptosis-related genes affected by TMZ. Notably, GADD45A was found to be up-regulated by TMZ in both cell cycle and apoptosis arrays. Furthermore, GADD45A knockdown (GADD45Akd) enhanced the cell growth arrest and cell death induced by TMZ, even in natural (T98) and adapted (TR-U373) TMZ-resistant cells. Interestingly, GADD45Akd decreased the expression of O6-methylguanine-DNA methyltransferase (MGMT) in TMZ-resistant cells (T98 and TR-U373). In MGMT-deficient/TMZ-sensitive cells (U87 and U373), GADD45Akd decreased TMZ-induced TP53 expression. Thus, in this study, we investigated the genes influenced by TMZ that were important in GBM therapy, and revealed that GADD45A plays a protective role against TMZ treatment which may through TP53-dependent and MGMT-dependent pathway in TMZ-sensitive and TMZ-resistant GBM, respectively. This protective role of GADD45A against TMZ treatment may provide a new therapeutic strategy for GBM treatment.

Early application of low-level laser may reduce the incidence of postherpetic neuralgia (PHN).

BACKGROUND: Postherpetic neuralgia (PHN) is difficult to treat, and currently there are no available treatments that effectively reduce its incidence. Low-level laser therapy (LLLT) has been proposed for indirect virus deactivation in treating recurrent herpes simplex infections. OBJECTIVE: This study seeks to investigate whether LLLT could reduce the incidence of PHN. METHODS: We retrospectively reviewed the incidence of PHN at the first, third, and sixth months after rash outbreak in 3 groups: the acute group of patients who received LLLT during the first 5 days; the subacute group of patients who received LLLT during days 6 to 14 of the eruption; and the control group of patients who did not receive LLLT. RESULTS: There were 48, 48, and 154 patients in the acute, subacute, and control groups, respectively. After adjusting for confounding factors, including age, sex, and use of famciclovir, the incidence of PHN was significantly lower in the acute group versus the control group after 1 month (odds ratio [OR] 0.21, P = .006, 95% confidence interval [CI] 0.068-0.632), 3 months (OR 0.112, P = .038, 95% CI 0.014-0.886), and 6 months (OR 0.123, P = .021, 95% CI 0-0.606). The subacute group only had a lower incidence (OR 0.187, P = .032, 95% CI 0.041-0.865) after 3 months when compared with the control group. LIMITATIONS: This is a retrospective study lacking double-blind randomization, and the placebo effect may be a major concern. Lack of standardized and prospective evaluation measures is also a limitation of this study. CONCLUSION: Applying LLLT within the first 5 days of herpes zoster eruption significantly reduced the incidence of PHN. LLLT may have the potential to prevent PHN, but further well-designed randomized controlled trials are required.

Efficacy of rituximab for pemphigus: a systematic review and meta-analysis of different regimens.

This meta-analysis examined the efficacy of different dosing regimens containing rituximab (RTX) in treating pemphigus. The analysis included 578 patients with pemphigus from 30 studies. Seventy-six percent of patients achieved complete remission (CR) after 1 cycle of RTX. Mean time to remission was 5.8 months, with a remission duration of 14.5 months and a 40% relapse rate. Eighteen patients (3.3%) developed serious adverse effects. The pooled estimate showed no significant differences in CR and relapse rates between patients treated with high-dose (near or ≥ 2,000 mg/cycle) vs. low-dose (< 1,500 mg/cycle) RTX. In the fully adjusted analysis, high-dose RTX was associated with longer duration of CR compared with low-dose RTX. No superiority of lymphoma protocol over rheumatoid arthritis or high-dose RTX over low-dose RTX was shown in other outcomes. RTX treatment is efficacious and well-tolerated in treating pemphigus. High-dose RTX treatment may lead to longer duration of remission. However, the choice of optimal regimen depends on the overall condition of the individual patient.