Search

Background: The performance of chest radiography-based age and sex prediction has not been well validated. We used a deep learning model to predict the age and sex of healthy adults based on chest radiographs (CXRs). Methods: In this retrospective study, 66,643 CXRs of 47,060 healthy adults were used for model training and testing. In total, 47,060 individuals (mean age ± standard deviation, 38.7 ± 11.9 years, 22,144 males) were included. By using chronological ages as references, mean absolute error (MAE), root mean square error (RMSE), and Pearson’s correlation coefficient were used to assess the model performance. Summarized class activation maps were used to highlight the activated anatomical regions. The area under the curve (AUC) was used to examine the validity for sex prediction. Results: When model predictions were compared with the chronological ages, the MAE was 2.1 years, RMSE was 2.8 years, and Pearson’s correlation coefficient was 0.97 (p &lt, 0.001). Cervical, thoracic spines, first ribs, aortic arch, heart, rib cage, and soft tissue of thorax and flank seemed to be the most crucial activated regions in the age prediction model. The sex prediction model demonstrated an AUC of &gt, 0.99. Conclusion: Deep learning can accurately estimate age and sex based on CXRs.

Background We aimed to examine the differences in the cost distributions, service use, and mortality outcomes, across major psychiatric disorders in Taiwan. Method A national cohort of adult patients (n = 68,068) who had newly received a diagnosis of schizophrenia, bipolar disorder, and major depressive disorder (MDD) was identified from the National Health Insurance Research Database and followed for the subsequent three years. Variations in the 1-year and 3-year healthcare cost distributions and mortality outcomes were examined according to age group (18–64 years, ≥65 years) and diagnosis. Results Regardless of age group, individuals with schizophrenia had the highest total and psychiatric healthcare costs. Healthcare costs for psychiatric services accounted for 84.25%, 60%, and 29.62% of the 1-year total healthcare costs for younger patients with a diagnosis of schizophrenia, bipolar disorder, and MDD, respectively. Psychiatric inpatient care costs constituted a major part of the 1-year psychiatric healthcare costs, e.g., 85.86% for schizophrenia patients aged 18–64 years, while psychiatric medication costs contributed to a relatively smaller part. For those older than 65 years, costs of other specialties for comorbid physical conditions were more prominent. Limitations The perspective of the current analysis was limited to healthcare services, and we were not able to analyse wider economic impacts. Conclusions Psychiatric inpatient care costs contributed to a significant share of psychiatric expenditures, emphasizing the need of developing strategies to reduce rehospitalisations. For those aged 65 years or older, efforts to improve interdisciplinary service care for comorbid physical conditions may be required.

Stenotic lesion rigidity (SLR) has an unclear influence on the outcome of percutaneous transluminal angioplasty and stenting (PTAS) for intracranial arterial stenosis. This study evaluated the outcome of PTAS and the relationship of vertebrobasilar SLR to features on vessel wall MRI (VW-MRI) for identifying pathologies of vertebrobasilar stenosis (VBS) and evaluating PTAS outcome. We retrospectively evaluated the results of PTAS in 31 patients with severe VBS. Stenotic lesions were classified as soft (based on predilatation pressure [PP] ≦ 4 atm) in 15 patients or hard (PP >4 atm) in 16 patients. We examined the relationship of SLR to clinical and MR findings. Patients with hard vs soft lesions had atherosclerosis (8/16 [50.0%] vs 2/15 [13.3%]), dissection (0/16 [0.0%] vs 12/15 [80.0%]), and dissection in atherosclerosis (8/16 [50.0%] vs 1/15 [6.7%], P 50%) in 5/15 (33.3%) vs 0/15 (0.0%) (P = 0.0421) in the 30 patients who successfully completed PTAS. Vertebrobasilar SLR correlated well with lesion etiology, findings on VW-MRI, and PTAS outcome. Patients with hard stenotic lesions need close follow-up after PTAS.

Background Whether the presence of hyperdense artery sign (HAS) correlates with clot characteristics and response to intravenous thrombotic therapy (IVT) remains to be determined. Given that the existent literature was mainly from the Western nations, the current study aimed to examine the relationships among HAS, clot characteristics, and outcome of IVT in a Han Chinese sample in Taiwan. The specific objectives are (1) to correlate HAS with clot characteristics and (2) to explore relationships between HAS and effectiveness/complication of IVT. Methods We enrolled 75 patients treated with IVT following acute large arterial infarctions. All patients had a baseline brain computed tomography and a follow-up image at 24 hours after thrombolysis. Correlations were explored between HAS and clot characteristics. Multivariable logistic regressions were employed to examine the relationships between HAS and response to IVT at 24 hours, including early improvement and all forms of intracerebral hemorrhage (ICH). Results In this Han Chinese sample in Taiwan, 50.7% of the patients had HAS and 64% had early improvements following IVT. Those with HAS tended to have clots lodging at main trunks of cerebral arteries but no significant associations were found between HAS and stroke etiology. In multivariable logistic regressions, HAS predicted neither early improvement nor ICH complication. Conclusions Instead of clot etiology, we found that it might be clot location that correlated with HAS. HAS was not associated with early improvement or ICH complication after IVT in this Han Chinese sample. We also showed that some other patient characteristics were likely to influence outcomes of IVT, which warrant clinical attention.

There is a lack of clarity in terms of cost-effectiveness and cost-utility comparisons across different outpatient (OPD) follow-up patterns in discharged patients with bipolar disorder (BD). In this study, adult patients hospitalised for BD treatment (n = 1,591) were identified from the National Health Insurance Research Database in Taiwan. With survival as the effectiveness measure and quality-adjusted life years (QALYs) as the utility measure, a cost-effectiveness and cost-utility analysis was conducted over the 3-year follow-up period by post-discharge frequency of OPD visits. Compared to those making 1–7, 8–12 and 18 or more OPD visits, BD patients making 13–17 OPD visits within the first year after discharge had the lowest psychiatric and total healthcare costs over the follow-up period. With survival status as the effectiveness outcome, making 13–17 OPD visits was more likely to be the cost-effective option, as revealed by incremental cost-effectiveness ratios. Cost-utility analysis demonstrated that having 13–17 OPD visits was probably the more cost-effective option when considering QALYs; for instance, if society was willing to pay NTD1.5 million for one additional QALY, there was a 75.2% (psychiatric costs) to 77.4% (total costs) likelihood that 13–17 OPD visits was the most cost-effective option. In conclusion, post-discharge OPD appointments with a frequency of 13–17 visits within the first year were associated with lower psychiatric and total healthcare costs in the subsequent 3 years. Having an adequate outpatient follow-up frequency was likely to be cost-effective in the management of discharged patients with BD in this real-world setting.

Background Although depression and chronic pain frequently co-occur, there is a lack of clarity in the literature regarding the cost-effectiveness and cost-utility of antidepressants in the presence of these two conditions. From the perspective of healthcare provider, the current study aims to compare the cost-effectiveness and cost-utility of antidepressants in a national cohort of depressed patients with and without comorbid pain conditions. Methods Adult patients prescribed with antidepressants for depression were identified from the National Health Insurance Research Database in Taiwan (n = 96,501). By using remission as effectiveness measure and quality-adjusted life years (QALYs) as utility measure, the cost-effectiveness and cost-utility were compared across selective serotonin reuptake inhibitors (SSRIs), serotonin norepinephrine reuptake inhibitors (SNRIs), and tricyclic antidepressants (TCAs), as well as by the presence of comorbid painful physical symptoms (PPS). Results SSRIs dominated SNRIs in both the cost-effectiveness and cost-utility regardless of comorbid PPS. In comparison with TCAs, SSRIs were likely to be the cost-effective option for patients without PPS. In patients with PPS, the cost-utility advantage for SSRIs over TCAs varied with threshold willingness-to-pay levels. Comorbid PPS may be considered an effect modifier of the cost-utility comparisons between SSRIs and TCAs. Conclusions For depressed patients without PPS, SSRIs are likely to be cost-effective in improving remission rates and QALYs compared to TCAs and SNRIs. However, to improve cost-utility in those with comorbid PPS, people need to choose between SSRIs and TCAs according to threshold willingness-to-pay levels. Future research is warranted to clarify the impacts of different pain conditions on the economic evaluations of pharmacological treatments in patients with depression.

Objective There is a lack of clarity in the literature regarding the cost-effectiveness and cost-utility of antidepressants for treating real-world patients. The impact of comorbid cardiovascular disease (CVD) on the economic evaluations of antidepressants remains to be determined. Method Adult patients prescribed with antidepressants for depression were identified from the National Health Insurance Research Database in Taiwan. A cost-effectiveness and cost-utility analysis was conducted comparing selective serotonin reuptake inhibitors (SSRIs), serotonin norepinephrine reuptake inhibitors (SNRIs), and tricyclic antidepressants (TCAs), and by the presence of comorbid CVD. Results In terms of treatment success rates, SSRIs were the most cost-effective option compared to TCAs and SNRIs as revealed in the incremental cost-effectiveness ratios. The cost-effectiveness acceptability curves further showed differential findings in the cost-utility results by the presence of comorbid CVD. Conclusion To improve treatment success rates and quality-adjusted life years, SSRIs can be considered the most cost-effective option. Future research is needed to further clarify the impacts of physical comorbidities and other associated factors on the cost-effectiveness and cost-utility of pharmacological treatments in patients with depression.

Background Depression and headache are highly prevalent in clinical settings. The co-occurrence of headache may impact choice of antidepressants, healthcare utilisation, and outcomes in patients with depression. The current study aims to examine the cost-effectiveness and cost-utility of different antidepressants for treating patients with depression and comorbid headache disorders. Methods Adult patients prescribed with antidepressants for depression ( n =96,501) were identified from the National Health Insurance Research Database in Taiwan. A cost-effectiveness and cost-utility analysis was conducted comparing selective serotonin reuptake inhibitors (SSRIs), serotonin norepinephrine reuptake inhibitors (SNRIs), and tricyclic antidepressants (TCAs), and by the presence of comorbid headache disorders and other pain conditions. Results In this study, SSRIs dominated SNRIs in both cost-effectiveness and cost-utility. As revealed in the cost-effectiveness acceptability curves, TCAs were likely to have a cost-utility advantage compared to SSRIs and SNRIs in improving quality-adjusted life years (QALYs) for patients with comorbid headache; SSRIs remained as the most cost-effective option for patients with other pain conditions. Limitations Limitations include the use of proxy definition of remission as effectiveness measure and the adoption of utility values from previous studies. Conclusions Given a pre-determined willingness-to-pay level, TCAs can be considered as a cost-effective option to improve QALYs for depressed patients with headache disorders. Future research is needed to further clarify factors influencing the cost-effectiveness and cost-utility of pharmacological treatments in depressed patients with specific pain conditions.

Background Thunderclap headache (TCH) is a sudden headache (SH) with accepted criteria of severe intensity and onset to peak within one minute. It is a well-known presentation for subarachnoid hemorrhage (SAH) but most patients with TCH or SH run a benign course without identifiable causes. Reversible cerebral vasoconstriction syndrome (RCVS), a recently recognized syndrome characterized by recurrent TCH attacks, has been proposed to account for most of these patients. Methods We recruited consecutive patients presenting with SH at our headache clinic. Computed tomography and/or magnetic resonance imaging with angiography were performed to exclude structural causes and to identify vasoconstriction. Catheter angiography and lumbar puncture were performed with patients consent. Reversibility of vasoconstriction was confirmed by follow-up study. Results From July 2010 to June 2013, 31 patients with SH were recruited. Twenty-four (72.7%) of these SH patients exhibited headache fulfilling the TCH criteria. The diagnosis of RCVS was confirmed in 14 (45.2%) of patients with SH and 11 (45.8%) of patients with TCH. Other diagnoses were as follows: primary headaches (SH: 41.9%, TCH: 45.8%) and other secondary causes (SH: 12.9%, TCH: 8.3%). Compared with non-RCVS patients, patients with RCVS were older (50.8 ± 9.3 years vs. 40.8 ± 10.0 years, P = 0.006) and less likely to experience short headache duration of P = 0.007). Patients with RCVS were more likely to cite bathing (42.9% vs. 0%, P = 0.004) and less likely to cite exertion (0% vs. 29.4%, P = 0.048) as headache triggers. Conclusions Reversible cerebral vasoconstriction syndrome is a common cause of SH and TCH. Considering the potential mortality and morbidity of RCVS, systemic examination of cerebral vessels should be performed in these patients.

Carotid-esophageal fistula (CEF) could be a serious complication of esophageal cancer in a patient receiving radiotherapy. We reported a 47-year-old male patient with advanced cervical esophageal cancer under radiotherapy who developed CEF with the presentations of unstable vital signs and disturbances of consciousness. Carotid-esophageal fistula-associated life-threatening conditions of carotid blowout syndrome and cerebral gas embolism were diagnosed after presentation. Subsequently, intramural dissection of esophageal and gastric walls, profound hemoperitoneum, and hypovolemic shock occurred. When a patient who had an underlying cervical esophageal cancer treated by radiotherapy develops unstable vital signs and neurological symptoms, CEF should be kept in mind in the differential diagnoses. Physicians must be alert of the associated complications of carotid blowout syndrome and cerebral gas embolism and perform timely management including decompression, fluid resuscitation, and aggressive endovascular procedure when indicated.

Artery of Percheron (AOP) is small perforating arteries supplying paramedian thalamus and mid-brain. The incidence of infarction is rare. We presented a 62-year-old man found conscious drowsy for 4 days. MRI revealed bilateral thalamic and midbrain infarction due to AOP occlusion.

Backgrounds The circle of Willis (CoW) is a primary collateral pathway that compensates quickly for a drop in cerebral blood flow. Using the complete CoW as a surrogate marker for good collateral circulation, its prognostic value after intravenous thrombolysis was examined. Methods We prospectively studied 64 consecutive patients with acute ischemic stroke treated with tissue plasminogen activator within 3 hours of stroke onset between October 2005 and June 2012 in our hospital. The study protocol was based on standard guidelines for intravenous thrombolysis. On computed tomographic angiography 24 hours after thrombolysis, the CoW was complete in 21 (32.8%) cases and incomplete in 43 (67.2%). Results Patients with complete CoW were more likely to have early improvement in National Institute of Health Stroke Scale (NIHSS) score (median improvement 2 vs 0 at 2 hours; 4 vs 1 at 24 hours), be independent at 3 months (42% vs 19%). In the incomplete CoW group, the rate of symptomatic intracerebral haemorrhage (SICH) according to the Safe Implementation of Thrombolysis in Stroke-Monitoring Study (SITS-MOST) definition was almost 3 times higher. Complete CoW was one of the strongest predictors of good functional outcome at 3 months (odds ratio 2.32; P = .01). Conclusions Complete CoW independently predicted functional independence and survival.