Predicting the COVID-19 mortality among Iranian patients using tree-based models: A cross-sectional study.

Background and Aims: To explore the use of different machine learning models in prediction of COVID-19 mortality in hospitalized patients. Materials and Methods: A total of 44,112 patients from six academic hospitals who were admitted for COVID-19 between March 2020 and August 2021 were included in this study. Variables were obtained from their electronic medical records. Random forest-recursive feature elimination was used to select key features. Decision tree, random forest, LightGBM, and XGBoost model were developed. Sensitivity, specificity, accuracy, F-1 score, and receiver operating characteristic (ROC)-AUC were used to compare the prediction performance of different models. Results: Random forest-recursive feature elimination selected following features to include in the prediction model: Age, sex, hypertension, malignancy, pneumonia, cardiac problem, cough, dyspnea, and respiratory system disease. XGBoost and LightGBM showed the best performance with an ROC-AUC of 0.83 [0.822-0.842] and 0.83 [0.816-0.837] and sensitivity of 0.77. Conclusion: XGBoost, LightGBM, and random forest have a relatively high predictive performance in prediction of mortality in COVID-19 patients and can be applied in hospital settings, however, future research are needed to externally confirm the validation of these models.

The functional impact of home-based self-rehabilitation following arthroscopic meniscus root repair.

BACKGROUND: Corona virus infectious pandemic makes outdoors rehabilitation a potential hazard. Patient education to perform simple home-based exercises seems to be an interesting and sometimes a mandatory option. This study provides a comparison between the conventional and home-based virtual rehabilitation after surgical repair of medial meniscus root tears. METHODS: In this prospective study, all patients who underwent medial meniscus posterior root repair with a modified trans-tibial pull-out technique from March 2019 to March 2021 were evaluated. Those who underwent surgery after December 2019 were trained to perform self-rehabilitation. The rest had undergone outdoors specialized rehabilitation according to a unified protocol and these were used as a historical control group. All patients were followed up for a minimum of 2 year after surgery. Final Lysholm scores were utilized to compare functional outcomes after considering the effect of age, body mass index and time from surgery by multivariate linear regression analysis. RESULTS: Forty-three consecutive patients with medial meniscal root tears were studied. Thirty-nine (90.7%) were women and 4 (9.3%) were men. The mean age of participants was 53.2 ± 8.1 years. The total Lysholm knee score, and all its items were significantly improved in both groups at a two-year follow-up (p < 0.05), except the "Using cane or crutches" item (p = 0.065). Nevertheless, the final Lysholm knee score improvement was higher in patients who performed outdoors specialized rehabilitation and in patients with shorter time-to-surgery. CONCLUSION: Regardless of age and gender, home-based rehabilitation after meniscal root repair with the modified trans-tibial pull-out technique improved the patients' function at a two-year follow-up. Nonetheless, this effect was still significantly lower than that of the outdoors specialized rehabilitation. Future work is required to clarify basic protocols for home-based tele-rehabilitation programs and determine clinical, radiological and functional results. LEVEL OF EVIDENCE: Level IV, therapeutic, historically controlled study.

A systematic review and meta-analysis of the epidemiological characteristics of COVID-19 in children.

BACKGROUND: Several individual studies from specific countries have reported rising numbers of pediatric COVID-19 cases with inconsistent reports on the clinical symptoms including respiratory and gastrointestinal symptoms as well as diverse reports on the mean age and household exposure in children. The epidemiological characteristics of COVID-19 in children are not fully understood, hence, comprehensive meta-analyses are needed to provide a better understanding of these characteristics. METHODS: This review was conducted in Medline, Scopus, Cochrane library, Embase, Web of Science, and published reports on COVID-19 in children. Data were extracted by two independent researchers and a third researcher resolved disputes. STATA software and the random-effect model were used in the synthesis of our data. For each model, the heterogeneity between studies was estimated using the Q Cochrane test. Heterogeneity and publication bias were calculated using the I2 statistic and Egger's/Begg's tests. RESULTS: The qualitative systematic review was performed on 32 articles. Furthermore, the meta-analysis estimated an overall rate of involvement at 12% (95% CI: 9-15%) among children, with an I2 of 98.36%. The proportion of household exposure was calculated to be 50.99% (95% CI: 20.80%-80.80%) and the proportion of admitted cases was calculated to be 45% (95% CI: 24%-67%). Additionally, the prevalence of cough, fatigue, fever and dyspnea was calculated to be 25% (95% CI: 0.16-0.36), 9% (95% CI: 0.03-0.18), 33% (95% CI: 0.21-0.47) and 9% (95% CI: 0.04-0.15), respectively. It is estimated that 4% (95% CI: 1-8%) of cases required intensive care unit admission. CONCLUSIONS: The pediatric clinical picture of COVID-19 is not simply a classic respiratory infection, but unusual presentations have been reported. Given the high incidence of household transmission and atypical clinical presentation in children, we strongly recommend their inclusion in research and population-based preventive measures like vaccination as well as clinical trials to ensure efficacy, safety, and tolerability in this age group.

Epidemiological pattern of pediatric trauma in COVID-19 outbreak: Data from a tertiary trauma center in Iran.

INTRODUCTION: In Iran, like most other countries, COVID-19 has had a deep impact on children's lives. Our hypothesis was that, a significant change in the number of pediatric injuries has happened in trauma centers. In the current study, we intend to identify the possible epidemiological shift in pediatric fracture patterns, by comparing the data from 'COVID-19 era' and the mean data from the past 2 years. To the best of our knowledge there are only few reports on epidemiology of pediatric fractures during the COVID-19 outbreak. METHODS: Data are reported in two sections. In the descriptive section, epidemiological data regarding pediatric fractures referred to Taleghani tertiary trauma center, including demographics, distribution curves, etiologies and fracture types are presented during the 'COVID era', from 1 March 2020 to 15 April 2020. In the comparative section, the aforementioned data are compared with mean data from the past 2 years, the 'non-COVID era'. RESULTS: Altogether 117 of the 288 trauma children (40.62%) had a fractured bone (145 fractures). Patients were mostly boys, with a mean age of 9.87 years (SD=5.27). The three most common fracture types in children included distal radius, mid-forearm and humeral supracondylar fractures. Compared to non-COVID era, the number of pediatric trauma admissions dropped from 589 to 288. No significant change happened in the mean age, male/female ratio and percentage of motor vehicle accidents. Proportion of proximal humeral, proximal forearm, carpal, and hand fractures declined. The number of open fractures significantly dropped (from 12 to 2). CONCLUSIONS: In Iran, overall trend of pediatric trauma has been decreasing during the outbreak; but the lack of reduction in proportion of accidents may pose an alarm that an effective lock-down has not been imposed. This study has implications as to preparing appropriate resources particular to common "COVID era fractures".