Implementation Science is an interdisciplinary field dedicated to systematically studying how to effectively translate evidence-based research findings into practical application and implementation. In the health-related context, it focuses on enhancing the efficiency and quality of healthcare services, thereby facilitating the transition from scientific evidence to real-world practice. This article elaborates on Theories, Models, and Frameworks (TMF) within health-related Implementation Science, clarifying their basic concepts and classifications, and discussing their roles in guiding implementation processes. Furthermore, it reviews and prospects current research from three aspects: the constituent elements of TMF, their practical applications, and future directions. Five representative frameworks are emphasized, including the Consolidated Framework for Implementation Research (CFIR), the Practical Robust Implementation and Sustainability Model (PRISM), the Exploration, Preparation, Implementation, Sustainment (EPIS)framework, the Behavior Change Wheel (BCW), and the Normalization Process Theory (NPT). Additionally, resources such as the Dissemination & Implementation Models Webtool and the T-CaST tool are introduced to assist researchers in selecting appropriate TMFs based on project-specific needs.

This study summarizes the construction background, rules and regulations and institutional settings of the MOOC and Micro-course Club in the Second Hospital of Jilin University, discusses the means of teacher training for clinical teachers, and shows the application effect of the club. At the same time, the related problems encountered in the process of club construction are summarized and reflected. The construction of MOOC and micro-course clubs is conducive to improving the information-based teaching level of clinical teachers, and also provides new inspiration and ideas for the construction of medical clubs.

Objective:To analyze the clinical features and magnetic resonance imaging of non-malignant patients assigned to Prostate Imaging Reporting And Data System (PI-RADS) 5 score.Methods:We performed a retrospective review of 289 patients who underwent magnetic resonance ultrasound targeted combined system biopsy with PI-RADS 5 lesions in the First Affiliated Hospital of Nanjing Medical University between May 2019 and July 2021. The median age 72 (66, 77)years, median body mass index 24.4(22.3, 27.1)kg/m 2, median prostate volume (PV) 37.39(29.39, 48.86) ml, median PSA 22.24(10.91, 62.69) ng/ml, and median PSAD 0.53(0.30, 1.52)ng/ml 2 were recorded. According to the biopsy pathological results, all patients were divided into benign lesion group and prostate cancer group. PSA, PSAD, PV, and apparent diffusion coefficient (ADC) values were compared, and magnetic resonance imaging and clinical characteristics of patients with biopsy benign lesions were analyzed. Results:There were 11 cases (3.8%) with benign lesion and 278 cases (96.2%) with prostate cancer. The characters of 11 negative biopsy cases were displayed as follows: median age 69(66, 79)years, median body mass index 22.0(21.0, 25.5)kg/m 2, median PV 62.90(38.48, 71.96)ml, median PSA 5.55(2.99, 20.52)ng/ml, median PSAD 0.16(0.07, 0.24) ng/ml 2, median ADC 714.47(701.91, 801.26)×10 -6 mm 2/s, abnormal digital rectal and amination in 5 cases, smoking in 7 cases, and alcohol consumption in 4 cases. The median PV [62.90(38.48, 71.96) vs. 37.21(29.22, 47.82)ml, P<0.01], the PSA value [5.55(2.99, 20.52) vs. 23.53(11.14, 65.98)ng/ml, P<0.01], and the PSAD value [0.16(0.07, 0.24) vs. 0.58(0.31, 1.57)ng/ml 2, P<0.01] were significantly different between benign condition group and prostate carcinoma group. Benign condition group included 5 chronic prostatitis, 2 acute prostatitis (1 with focal adenocarcinoma), 2 granulomatous inflammation, and 2 tuberculous granulomatous inflammation. In 7 benign cases, PSA was less than 10 ng/ml, combined with frequent urination, urgency of urination and incontinence were founded. In 8 benign cases, the area of lesion was more than 50% of the total prostate area in the axial position and the imaging of magnetic resonance were diffused, with regular shape and uniform signal. The imaging of symmetrical distribution was in 6 cases. Conclusions:The benign condition with PI-RADS 5 lesions included chronic prostatitis, acute prostatitis, granulomatous inflammation and tuberculous granulomatous inflammation, among which prostatitis was the most common cause. The PSA value were less than 10 ng/ml in most benign cases, with symptoms such as frequent urination, urgency of urination and incontinence. The imaging of magnetic resonance were diffused, symmetrically distributed, with regular shape and uniform signal.

Traumatic cerebrospinal fluid leakage commonly presents in traumatic brain injury patients, and it may lead to complications such as meningitis, ventriculitis, brain abscess, subdural hematoma or tension pneumocephalus. When misdiagnosed or inappropriately treated, traumatic cerebrospinal fluid leakage may result in severe complications and may be life-threatening. Some traumatic cerebrospinal fluid leakage has concealed manifestations and is prone to misdiagnosis. Due to different sites and mechanisms of trauma and degree of cerebrospinal fluid leak, treatments for traumatic cerebrospinal fluid leakage varies greatly. Hence, the Craniocerebral Trauma Professional Group of Neurosurgery Branch of Chinese Medical Association and the Neurological Injury Professional Group of Trauma Branch of Chinese Medical Association organized relevant experts to formulate the " Chinese expert consensus on the diagnosis and treatment of traumatic cerebrospinal fluid leakage in adults ( version 2023)" based on existing clinical evidence and experience. The consensus consisted of 16 recommendations, covering the leakage diagnosis, localization, treatments, and intracranial infection prevention, so as to standardize the diagnosis and treatment of traumatic cerebrospinal fluid leakage and improve the overall prognosis of the patients.

To compare the performance of generalized additive model (GAM) and long short-term memory recurrent neural network (LSTM-RNN) on the prediction of daily admissions of respiratory diseases with comorbid diabetes. Daily data on air pollutants, meteorological factors and hospital admissions for respiratory diseases from Jan 1st, 2014 to Dec 31st, 2019 in Beijing were collected. LSTM-RNN was used to predict the daily admissions of respiratory diseases with comorbid diabetes, and the results were compared with those of GAM. The evaluation indexes were calculated by five-fold cross validation. Compared with the GAM, the prediction errors of LSTM-RNN were significantly lower [root mean squared error (RMSE): 21.21±3.30 vs. 46.13±7.60, <0.01; mean absolute error (MAE): 14.64±1.99 vs. 36.08±6.20, <0.01], and the value was significantly higher (0.79±0.06 vs. 0.57±0.12, <0.01). In gender stratification, RMSE, MAE and values of LSTM-RNN were better than those of GAM in predicting female admission (all <0.05), but there were no significant difference in predicting male admission between two models (all >0.05). In seasonal stratification, RMSE and MAE of LSTM-RNN were lower than those of GAM in predicting warm season admission (all <0.05), but there was no significant difference in value (>0.05). There were no significant difference in RMSE, MAE and between the two models in predicting cold season admission (all >0.05). In the stratification of functional areas, the RMSE, MAE and values of LSTM-RNN were better than those of GAM in predicting core area admission (all <0.05). has lower prediction errors and better fitting than the GAM, which can provide scientific basis for precise allocation of medical resources in polluted weather in advance.
Beijing/epidemiology* ; Diabetes Mellitus/epidemiology* ; Female ; Hospitalization ; Humans ; Male ; Memory, Short-Term ; Neural Networks, Computer

Objective:To explore the predictive value of transvaginal ultrasound measurement of cervical length (CL) in the first and second trimester on spontaneous preterm birth in singleton pregnant women.Methods:This study retrospectively recruited 2 254 singleton pregnancies without severe comorbidities at Peking University First Hospital from January 2019 to June 2019. CL was measured for all subjects using transvaginal ultrasound in the first (11-13 +6 weeks) and second trimester (21-23 +6 weeks). Differences in CL between women with preterm (preterm group) and full-term delivery (full-term group) as well as the CL during the first and second trimester were compared. The independent risk factors for preterm birth and the predictive value of CL in the first and second trimester for spontaneous preterm birth were also explored. Fisher's exact test, t-test, χ2 test, and logistic regression analysis, etc, were adopted for statistical analysis. Results:(1) For the 2 254 subjects, CL measured in the first trimester and second trimester were (36.1±4.2) mm (22.4-52.6 mm) and (36.9±5.3) mm (2.9-59.7 mm), respectively. The incidence of short cervix in the first trimester and second trimester were 0.31% (7/2 254) and 1.46% (33/2 254), respectively. When CL was ≤25.0 mm ( OR=43.92, 95% CI:6.83-282.49) or >25.0-≤30.3 mm ( OR=6.59, 95% CI:1.97-22.0) in the first trimester, the risk of short cervix increased in the second trimester (both P<0.05). (2) The total incidence of preterm delivery was 3.06% (69/2 254). CL and the incidence of short cervix did not differ significantly in the first trimester between the preterm and full-term group [(35.2±4.5) and (36.1±4.1) mm, t=－1.78, P=0.076; 1.5% (1/69) and 0.3% (6/2 185), χ 2=2.98, P=0.084]. Compared with the full-term group, CL was shorter and the incidence of short cervix was higher in the second trimester in the preterm group [(33.6±6.7) vs (37.0±5.2) mm, t=－5.12;8.7% (6/69) vs 1.2% (27/2 185), χ 2=25.80, P<0.001]. (3) Multivariate regression analysis showed that age ≥35 years ( OR=2.05, 95% CI:1.22-3.46), history of spontaneous preterm birth ( OR=25.25, 95% CI:5.01-127.28), conception assisted by reproductive technology ( OR=10.39, 95% CI:2.39-50.33), and short cervix during the second trimester were independent risk factors for premature delivery. (4) There was no significant difference in the risk of preterm delivery when comparing to those with CL≤25.0 mm, >25.0-≤30.3 mm, >30.3-≤33.0 mm, >33.0-≤35.7 mm, >35.7-≤38.7 mm women with CL>38.7 mm during the first trimester (all P>0.05). The risk of premature delivery was relatively increased for those with CL≤25.0 mm,>25.0-≤29.5 mm, >29.5-≤33.6 mm, >33.6~≤36.8 mm, >36.8~≤40.1 mm during the second trimester compared to those with CL>40.1 mm [ OR (95% CI):17.64 (4.99-62.32), 6.89 (2.11-22.55), 3.58 (1.34-9.59), 4.04 (1.58-10.32), 3.34 (1.28-8.67), respectively , all P<0.05]. (5) When CL≤25.0 mm and ≤29.5 mm in the second trimester were used as the cut-off value, the prediction of preterm delivery was with a sensitivity of 8.70% and 17.39%, specificity of 98.80% and 95.29%, positive predictive value of 18.20% and 10.43%, negative predictive value of 97.16% and 97.34%, and the accuracy rate of 96.01% and 92.90%, respectively. Conclusions:There were no significant differences in CL and the incidence of short cervix during the first trimester among women with preterm or full-term delivery. CL in the first trimester is not an independent risk factor for preterm birth, but the risk of short cervix in the second trimester is increased when CL≤30.3 mm in the first trimester. The shorter the cervix during the second trimester, the greater the risk of preterm birth.

Objective:To provide consistent data basis for the application of reference intervals for children blood cell analysis in different testing systems.Methods:According to the requirements of American Institute for Clinical and Laboratory Standardization (CLSI) EP9-A3 document, 45 samples were collected and Sysmex XN20-A1 were used as reference system. Beckman DxH800, Siemens ADVIA 2120i, and Mindray BC5310 were comparison systems. Complete blood count and leukocyte classification were performed by four systems. The outliers of the detection results were tested by the generalized extreme student deviate (ESD) method. An optimal regression model was selected by scatter diagram, deviation diagram and frequency distribution diagram, which was used to fit the regression equation and calculate the deviation at the medical decision level and reference interval. The acceptable range for blood count deviation was cited from the Analytical Quality Specifications for Routine Tests in Clinical Hematology. The acceptable range for leukocyte classification was based on the EQA program of Royal College of Pathologists of Australasia (RCPA).Results:After the outliers were deleted, the scatter plot showed a linear relationship between the reference system and the three comparison systems. The deviation plot showed that the differences were variable. Deming regression or Passing-Bablok regression was selected according to the data distribution. The determination coefficient R2 of reference system and three comparison systems ranged from 0.95 to 0.99 in blood count and leukocyte classification. At the upper and lower limits of the reference interval, the deviations between XN-20A1 and ADVIA 2120 system were all acceptable, except for MONO# at 0.12×10 9/L. The deviations of all parameters at medical decision level were within acceptable ranges. The lower limit of PLT is partially unacceptable at the level of medical decision related to treatment and prognosis. Conclusions:The results of complete blood count and leukocyte classification in reference system and the comparison system had good consistency within the children′s reference interval. Our study provided a scientific basis for the feasibility of adopting a unified reference interval for different detection systems.

Objective:This study aimed to observe whether the treatment-free remission (TFR) of second-generation tyrosine kinase inhibitors (TKI) in chronic myeloid leukemia (CML) is better than imatinib (IM) .Methods:The clinical data of 274 CML patients who discontinued treatment and with complete clinical data were retrospectively studied from June 2013 to March 2021. Using both univariate and multivariate Cox proportional hazards regression models, risk factors influencing TFR outcomes after drug withdrawal in CML patients were assessed.Results:A total of 274 patients were enrolled, 140 patients were women (51.1%) , with a median age of 48 (9-84) years at the time of TKI discontinuation. Prior to TKI discontinuation, 172 (62.8%) patients were treated with IM, and 102 (37.2%) had received second-generation TKI treatment, including 73 patients who had shifted from IM to a second-generation TKI and 29 patients who used second-generation TKI as the first-line treatment. The rationale for converting to a second-generation TKI are as follows: 37 patients aimed deep molecular response (DMR) to achieve TFR, seven patients changed due to IM intolerance, and 29 patients changed because of failure to achieve the optimal treatment response. The use of the last type of TKI included 96 patients (94.1%) with nilotinib, three patients (2.9%) with dasatinib, and two patients (2%) with flumatinib, including one patient who changed to IM due to second-generation TKI intolerance. No statistical differences were found in the median age at diagnosis and TKI discontinuation, sex, Sokal score, IFN treatment before TKI, median time of TKI treatment to achieve DMR, and the reasons for TKI discontinuation between the second TKI and IM ( P>0.05) .The median cumulative treatment time of TKI (71.5 months vs 88 months, P<0.001) , the last TKI median treatment time (60 months vs 88 months, P<0.001) , and the median duration of DMR (58 months vs 66 months, P=0.002) were significantly shorter in the second-generation TKI compared with IM. In the median follow-up of 22 (6-118) months after TKI discontinuation, 88 patients (32.1%) had lost their MMR at a median of 6 (1-91) months; of the 53 patients (60.2%) who lost MMR within 6 months, the overall TFR rate was 67.9%, and the cumulative TFR rates at 12 and 24 months were 70.5% and 67.5%, respectively. Withdrawal syndrome occurred in 26 patients (9.5%) . For patients who restarted TKI treatment, 72 patients (83.7%) achieved DMR again at a median treatment of 4 (1 to 18) months. The univariate analysis showed that the TFR rate of patients treated with second-generation TKI was significantly higher than those who were treated with IM (77.5% vs 62.2%, P=0.041) . A further subgroup analysis found that the TFR rate of the second-generation TKI patients was significantly higher than those treated with IM (80.8% vs 62.2%, P=0.026) . No significant difference was found in the second-generation TKI used as the first line treatment compared with those who were treated with IM (69.0% vs 62.2%, P=0.599) . The multivariate analysis results showed that second-generation TKI treatment was an independent prognostic factor affecting TFR in patients who discontinued TKI ( RR=1.827, 95% CI 1.015-3.288, P=0.044) . Conclusion:In the clinical setting, more CML patients rapidly achieved TFR using second-generation TKI than IM treatment.

In order to overcome the shortcomings of high false positive rate and poor generalization in the detection of microcalcification clusters regions, this paper proposes a method combining discriminative deep belief networks (DDBNs) to automatically and quickly locate the regions of microcalcification clusters in mammograms. Firstly, the breast region was extracted and enhanced, and the enhanced breast region was segmented to overlapped sub-blocks. Then the sub-block was subjected to wavelet filtering. After that, DDBNs model for breast sub-block feature extraction and classification was constructed, and the pre-trained DDBNs was converted to deep neural networks (DNN) using a softmax classifier, and the network is fine-tuned by back propagation. Finally, the undetected mammogram was inputted to complete the location of suspicious lesions. By experimentally verifying 105 mammograms with microcalcifications from the Digital Database for Screening Mammography (DDSM), the method obtained a true positive rate of 99.45% and a false positive rate of 1.89%, and it only took about 16 s to detect a 2 888 × 4 680 image. The experimental results showed that the algorithm of this paper effectively reduced the false positive rate while ensuring a high positive rate. The detection of calcification clusters was highly consistent with expert marks, which provides a new research idea for the automatic detection of microcalcification clusters area in mammograms.
Algorithms ; Breast Neoplasms/diagnostic imaging* ; Calcinosis/diagnostic imaging* ; Early Detection of Cancer ; Humans ; Mammography ; Neural Networks, Computer

Hospital logistics management provides critical support for clinical work, and the management of materials is key to logistics management. Based on an analysis of current logistics management of the hospital, desirable results on logistics materials management have been harvested. The measures taken include optimizing management workflow and reinforcing cost control, in combination with such efforts as regulations improvement, higher informatization level and staff teamwork building.