D-dimer Levels for the exclusion of pulmonary embolism: making sense of international guideline recommendations

Several international guidelines provide recommendations around the use of D-dimer testing for exclusion of pulmonary embolism, including the appropriate D-dimer threshold (or cutoff), but there is no consensus among them. We briefly discuss guideline variation, performance characteristics, and limitations of commercially available D-dimer assays in this setting, referencing the Clinical and Laboratory Standards Institute guidelines that recommend immunoassays with high sensitivity (≥97%) and negative predictive value (≥98%). While age-adjusted D-dimer and pretest-adjusted D-dimer are considered a safe strategy across predefined patient subgroups, clinicians need to recognize the different performance characteristics of D-dimer assays to enable safe clinical decisions for their patients. Importantly, D-dimer values must be correlated not only to clinical findings but also interpreted within the context of the accuracy and precision of the specific testing modality, adhering to manufacturer specifications that are approved by regulatory authorities.

Are Clinical Prediction Rules Used in Spinal Cord Injury Care? A Survey of Practice

Background

Accurate outcome prediction is desirable post spinal cord injury (SCI), reducing uncertainty for patients and supporting personalized treatments. Numerous attempts have been made to create clinical prediction rules that identify patients who are likely to recover function. It is unknown to what extent these rules are routinely used in clinical practice.

Objectives

To better understand knowledge of, and attitudes toward, clinical prediction rules amongst SCI clinicians in the United Kingdom.

Methods

An online survey was distributed via mailing lists of clinical special interest groups and relevant National Health Service Trusts. Respondents answered questions about their knowledge of existing clinical prediction rules and their general attitudes to using them. They also provided information about their level of experience with SCI patients.

Results

One hundred SCI clinicians completed the survey. The majority (71%) were unaware of clinical prediction rules for SCI; only 8% reported using them in clinical practice. Less experienced clinicians were less likely to be aware. Lack of familiarity with prediction rules was reported as being a barrier to their use. The importance of clinical expertise when making prognostic decisions was emphasized. All respondents reported interest in using clinical prediction rules in the future.

Conclusion

The results show widespread lack of awareness of clinical prediction rules amongst SCI clinicians in the United Kingdom. However, clinicians were positive about the potential for clinical prediction rules to support decision-making. More focus should be directed toward refining current rules and improving dissemination within the SCI community.

Predictors of Mortality in Veterans with Amyotrophic Lateral Sclerosis: Respiratory Status and Speech Disorder at Presentation

BACKGROUND There is a lack of accurate models to predict amyotrophic lateral sclerosis (ALS) disease course and outcomes. As a result, risk assessment and counseling, the timing of interventions, and their stratification in clinical trials are difficult. This study aimed to evaluate the association between symptoms at presentation and mortality. MATERIAL AND METHODS A single veterans hospital reviewed the electronic records of 105 veterans with ALS who were periodically followed in our ALS clinic between 2010 and 2021. A survival decision tree (≤3 or >3 years) was generated based on the statistical median survival of our data. The variables known to influence survival when alive were compared to patients who died. RESULTS The (mean±SD) age at onset was 62±11 years, M/F ratio 101: 4, and 90% were non-Hispanic whites. The initial score for the Amyotrophic Lateral Sclerosis Functional Rating Scale-Revised (ALSFRS-R) was 31±8.3. Dysarthria and shortness of breath (SOB) were present on initial presentation in 52 (49.5%) and 32 (30.5%) patients, respectively. Deaths occurred in 80 (76.2%) patients during the study period. The main cause of death was respiratory disease (failure and pneumonia, n=43 53.75%). Patients survived for >3 years on initial presentation with normal respiration and speech, compared to ≤3 years of survival in patients with dysarthria and SOB, irrespective of age. CONCLUSIONS This study suggests that for veterans with ALS, the main predictors of shorter survival were respiratory status and speech disorder on initial presentation to the clinic.

Machine learning-based prediction of vitamin D deficiency: NHANES 2001-2018

Background

Vitamin D deficiency is strongly associated with the development of several diseases. In the current context of a global pandemic of vitamin D deficiency, it is critical to identify people at high risk of vitamin D deficiency. There are no prediction tools for predicting the risk of vitamin D deficiency in the general community population, and this study aims to use machine learning to predict the risk of vitamin D deficiency using data that can be obtained through simple interviews in the community.

Methods

The National Health and Nutrition Examination Survey 2001-2018 dataset is used for the analysis which is randomly divided into training and validation sets in the ratio of 70:30. GBM, LR, NNet, RF, SVM, XGBoost methods are used to construct the models and their performance is evaluated. The best performed model was interpreted using the SHAP value and further development of the online web calculator.

Results

There were 62,919 participants enrolled in the study, and all participants included in the study were 2 years old and above, of which 20,204 (32.1%) participants had vitamin D deficiency. The models constructed by each method were evaluated using AUC as the primary evaluation statistic and ACC, PPV, NPV, SEN, SPE, F1 score, MCC, Kappa, and Brier score as secondary evaluation statistics. Finally, the XGBoost-based model has the best and near-perfect performance. The summary plot of SHAP values shows that the top three important features for this model are race, age, and BMI. An online web calculator based on this model can easily and quickly predict the risk of vitamin D deficiency.

Conclusion

In this study, the XGBoost-based prediction tool performs flawlessly and is highly accurate in predicting the risk of vitamin D deficiency in community populations.

Patient outcomes following medical emergency team review on general wards: Development of predictive models

AIM

To develop and internally validate risk prediction models for subsequent clinical deterioration, unplanned ICU admission and death among ward patients following medical emergency team (MET) review.

DESIGN

A retrospective cohort study of 1500 patients who remained on a general ward following MET review at an Australian quaternary hospital.

METHOD

Logistic regression was used to model (1) subsequent MET review within 48 h, (2) unplanned ICU admission within 48 h and (3) hospital mortality. Models included demographic, clinical and illness severity variables. Model performance was evaluated using discrimination and calibration with optimism-corrected bootstrapped estimates. Findings are reported using the TRIPOD guideline for multivariable prediction models for prognosis or diagnosis. There was no patient or public involvement in the development and conduct of this study.

RESULTS

Within 48 h of index MET review, 8.3% (n = 125) of patients had a subsequent MET review, 7.2% (n = 108) had an unplanned ICU admission and in-hospital mortality was 16% (n = 240). From clinically preselected predictors, models retained age, sex, comorbidity, resuscitation limitation, acuity-dependency profile, MET activation triggers and whether the patient was within 24 h of hospital admission, ICU discharge or surgery. Models for subsequent MET review, unplanned ICU admission, and death had adequate accuracy in development and bootstrapped validation samples.

CONCLUSION

Patients requiring MET review demonstrate complex clinical characteristics and the majority remain on the ward after review for deterioration. A risk score could be used to identify patients at risk of poor outcomes after MET review and support general ward clinical decision-making.

RELEVANCE TO CLINICAL PRACTICE

Our risk calculator estimates risk for patient outcomes following MET review using clinical data available at the bedside. Future validation and implementation could support evidence-informed team communication and patient placement decisions.

Prospective Validation of the Calgary Kids' Hand Rule: A Clinical Prediction Rule for Pediatric Hand Fracture Triage

Introduction: Pediatric hand fractures are common and routinely referred to surgeons, yet most heal well without surgical intervention. This trend inspired the development of the Calgary Kids' Hand Rule (CKHR), a clinical prediction rule designed to predict "complex" fractures that require surgical referral. The CKHR was adapted into a checklist whereby the presence of any 1 of 6 clinically or radiologically identifiable fracture characteristics predicts a complex fracture. The aim of this study was to assess the accuracy of the CKHR in a prospective sample of children with hand fractures. Methods: Physicians were asked to complete the CKHR checklist when referring pediatric patients (< 18 years) to hand surgeons at a Canadian pediatric hospital (April 2019-September 2020). Completed checklists represented predicted outcomes and were compared to observed outcomes (determined via chart review). Predictive accuracy (primary outcome) was evaluated based on sensitivity and specificity. Secondary outcomes were interrater reliability between referring physicians and surgeons, and survey assessment of CKHR user satisfaction. Results: In total 365 fractures were included, with only 16 requiring surgical intervention. Overall performance of the CKHR was good with 84% sensitivity and 71% specificity. Percent agreement between referring physicians and surgeons ranged from 84.1% to 96.3% on individual predictors, with 78.1% agreement on the presence of any predictors. Survey results showed general user satisfaction but also identified areas for improvement. Conclusion: This study posits the CKHR as an accurate and clinically useful prediction rule and highlights the importance of education for its effective use and eventual scale and spread.

Development and validation of a clinical prediction model for 90-day venous thromboembolism risk following total hip and total knee arthroplasty: a multinational study

BACKGROUND

The risk of venous thromboembolism (VTE) following total hip arthroplasty (THA) and total knee arthroplasty (TKA) is 1.0% to 1.5%, despite uniform thromboprophylaxis.

OBJECTIVES

To develop and validate a prediction model for 90-day VTE risk.

METHODS

A multinational cohort study was performed. For model development, records were used from the Oxford Royal College of General Practitioners Research and Surveillance Centre linked to Hospital Episode Statistics and Office of National Statistics UK routine data. For external validation, data were used from the Danish Hip and Knee Arthroplasty Registry, the National Patient Registry, and the National Prescription Registry. Binary multivariable logistic regression techniques were used for development.

RESULTS

In the UK data set, 64 032 THA/TKA procedures were performed and 1.4% developed VTE. The prediction model consisted of age, body mass index, sex, cystitis within 1 year before surgery, history of phlebitis, history of VTE, presence of varicose veins, presence of asthma, history of transient ischemic attack, history of myocardial infarction, presence of hypertension and THA or TKA. The area under the curve of the model was 0.65 (95% CI, 0.63-0.67). Furthermore, 36 169 procedures were performed in the Danish cohort, of whom 1.0% developed VTE. Here, the area under the curve was 0.64 (95% CI, 0.61-0.67). The calibration slope was 0.92 in the validation study and 1.00 in the development study.

CONCLUSION

This clinical prediction model for 90-day VTE risk following THA and TKA performed well in both development and validation data. This model can be used to estimate an individual's risk for VTE following THA/TKA.

Evaluating the diagnostic accuracy of a screening tool for low physical activity in independently ambulating adults with knee osteoarthritis: A prospective cohort study

BACKGROUND AND PURPOSE

Patients with knee osteoarthritis (OA) can benefit from resistance training exercises, range of motion and flexibility maintenance, and low-load aerobic exercises, as per the relevant clinical guidelines. However, certain patients might be unable to progress to higher physical levels despite such physical therapy programs. This study aimed to evaluate the diagnostic accuracy of a screening tool for determining physical activity levels in individuals with OA undergoing standard physical therapy regularly, using likelihood ratios and predictive values.

METHOD

This prospective observational study included 135 patients undergoing standard physical therapy for OA from six medical facilities. The primary outcome was low physical activity or moderate to high physical activity levels based on 1-month Self-Rating Frenchay activities index scores. Backward elimination was used to perform binomial logistic regression analysis after identifying the independent variables in a univariate logistic regression analysis. Among the independent variables adopted in the logistic regression model, receiver operating characteristic analysis using Youden's index was performed for quantitative variables, which were converted to binary values at the cut-off points. Subsequently, the clinical prediction rule (CPR) was derived.

RESULTS

According to the binomial logistic regression analysis, age, knee flexion muscle strength, and visual analog scale (VAS) were risk factors for low physical activity, and the CPR was derived from these variables. The pre-test probability of the low physical activity group was 37.0% (50 out of 135 participants). For a total CPR score of three points (one point for each item: age ≤69 years, knee flexion muscle strength ≤0.36 Nm/kg, and VAS ≥33 mm), the positive likelihood ratio was 13.60 and the post-test probability increased to 88.9%.

DISCUSSION

The CPR identified patients who might not benefit from the standard physical therapy program. This screening tool could improve patient management, allowing for more tailored approaches in physical therapy programs.

Diagnostic performance of the classic symptom "abdominal pain before vomiting" for pediatric acute appendicitis

BACKGROUND

Acute appendicitis is the most common type of acute abdomen that requires surgical intervention in children. According to general pediatric textbooks, the presence of vomiting before abdominal pain is considered a classic patient history item for excluding acute appendicitis. However, its diagnostic performance in the pediatric population has yet to be investigated.

METHODS

This was a single-center retrospective observational study involving 134 children who were admitted to the hospital with both abdominal pain and vomiting. The reference standard for appendicitis was defined by computed tomography scanning. The diagnostic performance of "abdominal pain before vomiting" was calculated and compared to those of the Alvarado score and pediatric appendicitis score.

RESULTS

The diagnostic performance of "abdominal pain before vomiting" was as follows: sensitivity of 75.3% (95% confidence interval [CI], 64.7-83.6), specificity of 25.0% (95% CI, 15.5-36.7), positive likelihood ratio of 1.00 (95% CI, 0.82-1.22), negative likelihood ratio of 0.99 (95% CI, 0.54-1.79), and diagnostic odds ratio of 1.02 (95% CI, 0.46-2.25). In contrast, the Alvarado score and pediatric appendicitis score (with a threshold of 4 points) demonstrated favorable sensitivity (98.3% [95% CI, 92.4-99.6]), negative predictive value (94.6% [95% CI, 78.4-98.8]), negative likelihood ratio (0.04 [95% CI, 0.01-0.23]), and diagnostic odds ratio (49.9 [95% CI, 6.88-243.2]).

CONCLUSION

In this study, "abdominal pain before vomiting" was associated with poor diagnostic performance for ruling out acute pediatric appendicitis. Thus, the diagnosis of acute appendicitis in the pediatric population should be made based on existing validated scoring systems such as the Alvarado score and pediatric appendicitis score.

Pulmonary embolism risk stratification: external validation of the 4-level Clinical Pretest Probability Score (4PEPS)

Background

The 4-level clinical pretest probability score (4PEPS) was recently introduced as a clinical decision rule for the diagnosis of pulmonary embolism (PE). Based on the score, patients are classified into clinical pretest probability categories (c-PTP). The "very low" category aims at excluding PE without further testing; "low" and "moderate" categories require D-dimer testing with specific thresholds, while patients with a "high" pretest directly proceed to imaging.

Objectives

To provide further external validation of the 4PEPS model.

Methods

The 4PEPS was applied to a previously collected prospective database of 756 patients with clinically suspected PE enrolled from European emergency departments in 2002 to 2003. The safety threshold for the failure rate in our study was calculated at 1.95% based on a 26% prevalence of PE in our study, as per the International Society on Thrombosis and Haemostasis Scientific and Standardization Committee guidance.

Results

Patients were classified as follows: 90 (12%) in the very low c-PTP group, of whom 5 (5.6%; 95% CI, 2.4%-12.4%) had PE; 363 (49%) in the low c-PTP group, of whom 34 had PE (9.4%); 246 (34%) in the moderate c-PTP group, of whom 124 (50%) had PE; and 35 (5%) in the high c-PTP group of whom 30 (86%) had PE. Overall, the failure rate of the 4PEPS was 9/734 (1.2%; 95% CI, 0.59%-2.23%) Overall, 9 out of 734 patients (1.2%; 95% CI, 0.59%-2.23%) were diagnosed with PE despite a negative 4PEPS rule; 5 (5.6%) from the very low c-PTP group, 3 (1.4%) in the low c-PTP group, and 1 (3.2%) in the moderate c-PTP group.

Conclusion

We provide external validation data of the 4PEPS. In this high-prevalence cohort (26% prevalence), PE prevalence in the very low-risk group was higher than expected. A prospective validation study is needed before implementing the 4PEPS model in routine clinical practice.

Inpatient nurses' preferences and decisions with risk information visualization

OBJECTIVE

We examined the influence of 4 different risk information formats on inpatient nurses' preferences and decisions with an acute clinical deterioration decision-support system.

MATERIALS AND METHODS

We conducted a comparative usability evaluation in which participants provided responses to multiple user interface options in a simulated setting. We collected qualitative data using think aloud methods. We collected quantitative data by asking participants which action they would perform after each time point in 3 different patient scenarios.

RESULTS

More participants (n = 6) preferred the probability format over relative risk ratios (n = 2), absolute differences (n = 2), and number of persons out of 100 (n = 0). Participants liked average lines, having a trend graph to supplement the risk estimate, and consistent colors between trend graphs and possible actions. Participants did not like too much text information or the presence of confidence intervals. From a decision-making perspective, use of the probability format was associated with greater concordance in actions taken by participants compared to the other 3 risk information formats.

DISCUSSION

By focusing on nurses' preferences and decisions with several risk information display formats and collecting both qualitative and quantitative data, we have provided meaningful insights for the design of clinical decision-support systems containing complex quantitative information.

CONCLUSION

This study adds to our knowledge of presenting risk information to nurses within clinical decision-support systems. We encourage those developing risk-based systems for inpatient nurses to consider expressing risk in a probability format and include a graph (with average line) to display the patient's recent trends.

A prognostic model and pre-discharge predictors of post-COVID-19 syndrome after hospitalization for SARS-CoV-2 infection

Background

Post-COVID-19 syndrome (PCS) has been increasingly recognized as an emerging problem: 50% of patients report ongoing symptoms 1 year after acute infection, with most typical manifestations (fatigue, dyspnea, psychiatric and neurological symptoms) having potentially debilitating effect. Early identification of high-risk candidates for PCS development would facilitate the optimal use of resources directed to rehabilitation of COVID-19 convalescents.

Objective

To study the in-hospital clinical characteristics of COVID-19 survivors presenting with self-reported PCS at 3 months and to identify the early predictors of its development.

Methods

221 hospitalized COVID-19 patients underwent symptoms assessment, 6-min walk test, and echocardiography pre-discharge and at 1 month; presence of PCS was assessed 3 months after discharge. Unsupervised machine learning was used to build a SANN-based binary classification model of PCS development.

Results

PCS at 3 months has been detected in 75% patients. Higher symptoms level in the PCS group was not associated with worse physical functional recovery or significant echocardiographic changes. Despite identification of a set of pre-discharge predictors, inclusion of parameters obtained at 1 month proved necessary to obtain a high accuracy model of PCS development, with inputs list including age, sex, in-hospital levels of CRP, eGFR and need for oxygen supplementation, and level of post-exertional symptoms at 1 month after discharge (fatigue and dyspnea in 6MWT and MRC Dyspnea score).

Conclusion

Hospitalized COVID-19 survivors at 3 months were characterized by 75% prevalence of PCS, the development of which could be predicted with an 89% accuracy using the derived neural network-based classification model.

Risk stratification of acute pulmonary embolism

Approximately 5% of pulmonary embolism (PE) cases present with persistent hypotension, obstructive shock, or cardiac arrest. Given the high short-term mortality, management of high-risk PE cases focuses on immediate reperfusion therapies. Risk stratification of normotensive PE is important to identify patients with an elevated risk of hemodynamic collapse or an elevated risk of major bleeding. Risk stratification for short-term hemodynamic collapse includes assessment of physiological parameters, right heart dysfunction, and identification of comorbidities. Validated tools such as European Society of Cardiology guidelines and Bova score can identify normotensive patients with PE and an elevated risk of subsequent hemodynamic collapse. At present, we lack high-quality evidence to recommend one treatment over another (systemic thrombolysis, catheter-directed therapy, or anticoagulation with close monitoring) for patients at elevated risk of hemodynamic collapse. Newer, less well-validated scores such as BACS and PE-CH may help identify patients at a high risk of major bleeding following systemic thrombolysis. The PE-SARD score may identify those at risk of major anticoagulant-associated bleeding. Patients at low risk of short-term adverse outcomes can be considered for outpatient management. The simplified Pulmonary Embolism Severity Index score or Hestia criteria are safe decision aids when combined with physician global assessment of the need for hospitalization following the diagnosis of PE.

Clinical hypoxemia score for outpatient child pneumonia care lacking pulse oximetry in Africa and South Asia

Background

Pulse oximeters are not routinely available in outpatient clinics in low- and middle-income countries. We derived clinical scores to identify hypoxemic child pneumonia.

Methods

This was a retrospective pooled analysis of two outpatient datasets of 3-35 month olds with World Health Organization (WHO)-defined pneumonia in Bangladesh and Malawi. We constructed, internally validated, and compared fit & discrimination of four models predicting SpO2 < 93% and <90%: (1) Integrated Management of Childhood Illness guidelines, (2) WHO-composite guidelines, (3) Independent variable least absolute shrinkage and selection operator (LASSO); (4) Composite variable LASSO.

Results

12,712 observations were included. The independent and composite LASSO models discriminated moderately (both C-statistic 0.77) between children with a SpO2 < 93% and ≥94%; model predictive capacities remained moderate after adjusting for potential overfitting (C-statistic 0.74 and 0.75). The IMCI and WHO-composite models had poorer discrimination (C-statistic 0.56 and 0.68) and identified 20.6% and 56.8% of SpO2 < 93% cases. The highest score stratum of the independent and composite LASSO models identified 46.7% and 49.0% of SpO2 < 93% cases. Both LASSO models had similar performance for a SpO2 < 90%.

Conclusions

In the absence of pulse oximeters, both LASSO models better identified outpatient hypoxemic pneumonia cases than the WHO guidelines. Score external validation and implementation are needed.

Implementation of an emergency department back pain clinical management tool on the early diagnosis and testing of spinal epidural abscess

BACKGROUND

Spinal epidural abscess (SEA) is a rare, catastrophic condition for which diagnostic delays are common. Our national group develops evidence-based guidelines, known as clinical management tools (CMT), to reduce high-risk misdiagnoses. We study whether implementation of our back pain CMT improved SEA diagnostic timeliness and testing rates in the emergency department (ED).

METHODS

We conducted a retrospective observational study before and after implementation of a nontraumatic back pain CMT for SEA in a national group. Outcomes included diagnostic timeliness and test utilization. We used regression analysis to compare differences before (January 2016-June 2017) and after (January 2018-December 2019) with 95% confidence intervals (CIs) clustered by facility. We graphed monthly testing rates.

RESULTS

In 59 EDs, pre versus post periods included 141,273 (4.8%) versus 192,244 (4.5%) back pain visits and 188 versus 369 SEA visits, respectively. After implementation, SEA visits with prior related visits were unchanged (12.2% vs. 13.3%, difference +1.0%, 95% CI -4.5% to 6.5%). Mean number of days to diagnosis decreased but not significantly (15.2 days vs. 11.9 days, difference -3.3 days, 95% CI -7.1 to 0.6 days). Back pain visits receiving CT (13.7% vs. 21.1%, difference +7.3%, 95% CI 6.1% to 8.6%) and MRI (2.9% vs. 4.4%, difference +1.4%, 95% CI 1.0% to 1.9%) increased. Spine X-rays decreased (22.6% vs. 20.5%, difference 2.1%, 95% CI -4.3% to 0.1%). Back pain visits receiving erythrocyte sedimentation rate or C-reactive protein increased (1.9% vs. 3.5%, difference +1.6%, 95% CI 1.3% to 1.9%).

CONCLUSIONS

Back pain CMT implementation was associated with an increased rate of recommended imaging and laboratory testing in back pain. There was no associated reduction in the proportion of SEA cases with a related prior visit or time to SEA diagnosis.

Validation of a prediction model for post-chemotherapy fibrosis in nonseminoma patients

OBJECTIVE

To validate Vergouwe's prediction model using the Swedish and Norwegian Testicular Cancer Group (SWENOTECA) RETROP database and to define its clinical utility.

MATERIALS AND METHODS

Vergouwe's prediction model for benign histopathology in post-chemotherapy retroperitoneal lymph node dissection (PC-RPLND) uses the following variables: presence of teratoma in orchiectomy specimen; pre-chemotherapy level of alpha-fetoprotein; β-Human chorionic gonadotropin and lactate dehydrogenase; and lymph node size pre- and post-chemotherapy. Our validation cohort consisted of patients included in RETROP, a prospective population-based database of patients in Sweden and Norway with metastatic nonseminoma, who underwent PC-RPLND in the period 2007-2014. Discrimination and calibration analyses were used to validate Vergouwe's prediction model results. Calibration plots were created and a Hosmer-Lemeshow test was calculated. Clinical utility, expressed as opt-out net benefit (NBopt-out ), was analysed using decision curve analysis.

RESULTS

Overall, 284 patients were included in the analysis, of whom 130 (46%) had benign histology after PC-RPLND. Discrimination analysis showed good reproducibility, with an area under the receiver-operating characteristic curve (AUC) of 0.82 (95% confidence interval 0.77-0.87) compared to Vergouwe's prediction model (AUC between 0.77 and 0.84). Calibration was acceptable with no recalibration. Using a prediction threshold of 70% for benign histopathology, NBopt-out was 0.098. Using the model and this threshold, 61 patients would have been spared surgery. However, only 51 of 61 were correctly classified as benign.

CONCLUSIONS

The model was externally validated with good reproducibility. In a clinical setting, the model may identify patients with a high chance of benign histopathology, thereby sparing patients of surgery. However, meticulous follow-up is required.

Optimizing Clinical Decision Making with Decision Curve Analysis: Insights for Clinical Investigators

A large number of prediction models are published with the objective of allowing personalized decision making for diagnostic or prognostic purposes. Conventional statistical measures of discrimination, calibration, or other measures of model performance are not well-suited for directly and clearly assessing the clinical value of scores or biomarkers. Decision curve analysis is an increasingly popular technique used to assess the clinical utility of a prognostic or diagnostic score/rule, or even of a biomarker. Clinical utility is expressed as the net benefit, which represents the net balance of patients' benefits and harms and considers, implicitly, the consequences of clinical actions taken in response to a certain prediction score, rule, or biomarker. The net benefit is plotted against a range of possible exchange rates, representing the spectrum of possible patients' and clinicians' preferences. Decision curve analysis is a powerful tool for judging whether newly published or existing scores may truly benefit patients, and represents a significant advancement in improving transparent clinical decision making. This paper is meant to be an introduction to decision curve analysis and its interpretation for clinical investigators. Given the extensive advantages, we advocate applying decision curve analysis to all models intended for use in clinical practice.

Spontaneous physical functional recovery after hospitalization for COVID-19: insights from a 1 month follow-up and a model to predict poor trajectory

Background

Long COVID syndrome has emerged as a new global healthcare challenge, with impaired physical performance being a prominent debilitating factor. Cardiopulmonary rehabilitation is a mainstay of management of symptomatic post-COVID patients, and optimization of candidate selection might allow for more effective use of available resources.

Methods

In order to study the natural dynamics and to identify predictors of physical functional recovery following hospitalization for COVID-19, 6 min walk test was performed pre-discharge in 176 patients (40% hypertensive, 53% female, mean age 53.2 ± 13.5 years) with re-evaluation at 1 month.

Results

Six min walk distance and the reached percent of predicted distance (6MWD%) were suboptimal at both visits-396 ± 71 m (68.7 ± 12.4%) pre-discharge and 466 ± 65 m (81.8 ± 13.6%) at 1 month. Associated changes included significant oxygen desaturation (2.9 ± 2.5 and 2.3 ± 2.2%, respectively) and insufficient increment of heart rate during the test (24.9 ± 17.5 and 28.2 ± 12.0 bpm) that resulted in low reached percent of individual maximum heart rate (61.1 ± 8.1 and 64.3 ± 8.2%). Automatic clusterization of the study cohort by the 6MWD% changes has allowed to identify the subgroup of patients with poor "low base-low increment" trajectory of spontaneous post-discharge recovery that were characterized by younger age (38.2 ± 11.0 vs. 54.9 ± 12.1, p < 0.001) but more extensive pulmonary involvement by CT (43.7 ± 8.8 vs. 29.6 ± 19.4%, p = 0.029) and higher peak ESR values (36.5 ± 9.7 vs. 25.6 ± 12.8, p < 0.001). Predictors of poor recovery in multivariate logistic regression analysis included age, peak ESR, eGFR, percentage of pulmonary involvement by CT, need for in-hospital oxygen supplementation, SpO2 and mMRC dyspnea score pre-discharge, and history of hypertension.

Conclusion

COVID-19 survivors were characterized by decreased physical performance pre-discharge as assessed by the 6 min walk test and did not completely restore their functional status after 1 month of spontaneous recovery, with signs of altered blood oxygenation and dysautonomia contributing to the observed changes. Patients with poor "low base-low increment" trajectory of post-discharge recovery were characterized by younger age but more extensive pulmonary involvement and higher peak ESR values. Poor post-discharge recovery in the study cohort was predictable by the means of machine learning-based classification model that used age, history of hypertension, need for oxygen supplementation, and ESR as inputs.

Predicting length of stay for acute medical admissions in New Zealand: the MALICE score

Predicting length of stay (LoS) in hospital can help guide patient placement, facilitate rapid discharge and aid identification of patients at risk of prolonged stay, in whom early multidisciplinary intervention is warranted. We aimed to pilot the applicability of a modified decision aid (MALICE score) for predicting LoS for acute medical admissions at a New Zealand hospital. A prospective pilot study of 220 acute general medical admissions was performed. Clinical records were reviewed and MALICE scores were calculated for each patient and compared with LoS data using the Kruskal-Wallis H test. A statistically significant increase in LoS was seen with rising MALICE scores (H value 26.85, P < 0.001). MALICE scoring could be employed to guide patient placement and identify patients at risk of prolonged stays, though further study of bedside feasibility and applicability is required.

Use of Valsalva Maneuver to Detect Late-Onset Delayed Orthostatic Hypotension

BACKGROUND

Standard autonomic testing includes a 10-minute head-up tilt table test to detect orthostatic hypotension. Although this test can detect delayed orthostatic hypotension (dOH) between 3 and 10 minutes of standing, it cannot detect late-onset dOH after 10 minutes of standing.

METHODS

To determine whether Valsalva maneuver responses can identify patients who would require prolonged head-up tilt table test to diagnose late-onset dOH; patients with immediate orthostatic hypotension (onset <3 minutes; n=176), early-onset dOH (onset between 3 and 10 minutes; n=68), and late-onset dOH (onset >10 minutes; n=32) were retrospectively compared with controls (n=114) with normal head-up tilt table test and composite autonomic scoring scale score of 0.

RESULTS

Changes in baseline systolic blood pressure at late phase 2 (∆SBPVM2), heart rate difference between baseline and phase 3 (∆HRVM3), and Valsalva ratio were lower and pressure recovery time (PRT) at phase 4 was longer in late-onset dOH patients than in controls. Differences in PRT and ∆HRVM3 remained significant after correcting for age. A PRT ≥2.14 s and ∆HRVM3 ≤15 bpm distinguished late-onset dOH from age- and sex-matched controls. Patients with longer PRT (relative risk ratio, 2.189 [1.579-3.036]) and lower ∆HRVM3 (relative risk ratio, 0.897 [0.847-0.951]) were more likely to have late-onset dOH. Patients with longer PRT (relative risk ratio, 1.075 [1.012-1.133]) were more likely to have early-onset than late-onset dOH.

CONCLUSIONS

Long PRT and short ∆HRVM3 can help to identify patients who require prolonged head-up tilt table test to diagnose late-onset dOH.

Establishment of a simplified score for predicting risk during intrahospital transport of critical patients: A prospective cohort study

AIMS AND OBJECTIVES

To establish a simple score that enables nurses to quickly, conveniently and accurately identify patients whose condition may change during intrahospital transport.

BACKGROUND

Critically ill patients may experience various complications during intrahospital transport; therefore, it is important to predict their risk before they leave the emergency department. The existing scoring systems were not developed for this population.

DESIGN

A prospective cohort study.

METHODS

This study used convenience sampling and continuous enrolment from 1 January, 2019, to 30 June, 2021, and 584 critically ill patients were included. The collected data included vital signs and any condition change during transfer. The STROBE checklist was used.

RESULTS

The median age of the modelling group was 74 (62, 83) years; 93 (19.7%) patients were included in the changed group, and 379 (80.3%) were included in the stable group. The five independent model variables (respiration, pulse, oxygen saturation, systolic pressure and consciousness) were statistically significant (p < .05). The above model was simplified based on beta coefficient values, and each variable was assigned 1 point, for a total score of 0-5 points. The AUC of the simplified score in the modelling group was 0.724 (95% CI: 0.682-0.764); the AUC of the simplified score in the validation group (112 patients) was 0.657 (95% CI: 0.566-0.741).

CONCLUSIONS

This study preliminarily established a simplified scoring system for the prediction of risk during intrahospital transport from the emergency department to the intensive care unit. It provides emergency nursing staff with a simple assessment tool to quickly, conveniently and accurately identify a patient's transport risk.

RELEVANCE TO CLINICAL PRACTICE

This study suggested the importance of strengthening the evaluation of the status of critical patients before intrahospital transport, and a simple score was formed to guide emergency department nurses in evaluating patients.

The Clinical Predictive Score for Prehospital Large Vessel Occlusion Stroke: A Retrospective Cohort Study in the Asian Country

Background

Large vessel occlusive (LVO) stroke causes severe disabilities and occurs in more than 37% of strokes. Reperfusion therapy is the gold standard of treatment. Studies proved that endovascular thrombectomy (EVT) is more beneficial and decreases mortality. This study aimed to evaluate the factor associated with LVO stroke in an Asian population and to develop the scores to predict LVO in a prehospital setting. The score will hugely contribute to the future of stroke care in prehospital settings in the aspect of transferal suspected LVO stroke patients to appropriate EVT-capable stroke centers.

Methods

This study was a retrospective cohort study using an exploratory model at the emergency department of Ramathibodi Hospital, Bangkok, Thailand, between January 2018 and December 2020. We included the stroke patients aged >18 who visit ED and an available radiologic report representing LVO. Those whose stroke onset was >24 hours and no radiologic report were excluded. Multivariable logistic regression analysis developed the prediction model and score for LVO stroke.

Results

A total of 252 patients met the inclusion criteria; 61 cases (24%) had LVO stroke. Six independent factors were significantly predictive: comorbidity with atrial fibrillation, clinical hemineglect, gaze deviation, facial palsy, aphasia, and cerebellar sign abnormality. The predicted score had an accuracy of 92.5%. The LVO risk score was categorized into three groups: low risk (LVO score <3), moderate risk (LVO score 3-6), and high risk (LVO score >6). The positive likelihood ratio to predicting LVO stroke were 0.12 (95% CI 0.06-0.26), 2.33 (95% CI 1.53-3.53) and 45.40 (95% CI 11.16-184.78), respectively.

Conclusion

The Large Vessel Occlusion (LVO) Risk Score provides a screening tool for predicting LVO stroke. A clinical predictive score of ≥3 appears to be associated with LVO stroke.

US and Dutch Perspectives on the Use of COVID-19 Clinical Prediction Models: Findings from a Qualitative Analysis

INTRODUCTION

Clinical prediction models (CPMs) for coronavirus disease 2019 (COVID-19) may support clinical decision making, treatment, and communication. However, attitudes about using CPMs for COVID-19 decision making are unknown.

METHODS

Online focus groups and interviews were conducted among health care providers, survivors of COVID-19, and surrogates (i.e., loved ones/surrogate decision makers) in the United States and the Netherlands. Semistructured questions explored experiences about clinical decision making in COVID-19 care and facilitators and barriers for implementing CPMs.

RESULTS

In the United States, we conducted 4 online focus groups with 1) providers and 2) surrogates and survivors of COVID-19 between January 2021 and July 2021. In the Netherlands, we conducted 3 focus groups and 4 individual interviews with 1) providers and 2) surrogates and survivors of COVID-19 between May 2021 and July 2021. Providers expressed concern about CPM validity and the belief that patients may interpret CPM predictions as absolute. They described CPMs as potentially useful for resource allocation, triaging, education, and research. Several surrogates and people who had COVID-19 were not given prognostic estimates but believed this information would have supported and influenced their decision making. A limited number of participants felt the data would not have applied to them and that they or their loved ones may not have survived, as poor prognosis may have suggested withdrawal of treatment.

CONCLUSIONS

Many providers had reservations about using CPMs for people with COVID-19 due to concerns about CPM validity and patient-level interpretation of the outcome predictions. However, several people who survived COVID-19 and their surrogates indicated that they would have found this information useful for decision making. Therefore, information provision may be needed to improve provider-level comfort and patient and surrogate understanding of CPMs.

HIGHLIGHTS

While clinical prediction models (CPMs) may provide an objective means of assessing COVID-19 prognosis, provider concerns about CPM validity and the interpretation of CPM predictions may limit their clinical use.Providers felt that CPMs may be most useful for resource allocation, triage, research, or educational purposes for COVID-19.Several survivors of COVID-19 and their surrogates felt that CPMs would have been informative and may have aided them in making COVID-19 treatment decisions, while others felt the data would not have applied to them.

Do Thresholds for Invasive Ventilation in Hypoxemic Respiratory Failure Exist? A Cohort Study

Rationale: Invasive ventilation is a significant event for patients with respiratory failure. Physiologic thresholds standardize the use of invasive ventilation in clinical trials, but it is unknown whether thresholds prompt invasive ventilation in clinical practice. Objectives: To measure, in patients with hypoxemic respiratory failure, the probability of invasive ventilation within 3 hours after meeting physiologic thresholds. Methods: We studied patients admitted to intensive care receiving Fi_O2 of 0.4 or more via nonrebreather mask, noninvasive positive pressure ventilation, or high-flow nasal cannula, using data from the Medical Information Mart for Intensive Care (MIMIC)-IV database (2008-2019) and the Amsterdam University Medical Centers Database (AmsterdamUMCdb) (2003-2016). We evaluated 17 thresholds, including the ratio of arterial to inspired oxygen, the ratio of saturation to inspired oxygen ratio, composite scores, and criteria from randomized trials. We report the probability of invasive ventilation within 3 hours of meeting each threshold and its association with covariates using odds ratios (ORs) and 95% credible intervals (CrIs). Measurements and Main Results: We studied 4,726 patients (3,365 from MIMIC, 1,361 from AmsterdamUMCdb). Invasive ventilation occurred in 28% (1,320). In MIMIC, the highest probability of invasive ventilation within 3 hours of meeting a threshold was 20%, after meeting prespecified neurologic or respiratory criteria while on vasopressors, and 19%, after a ratio of arterial to inspired oxygen of <80 mm Hg. In AmsterdamUMCdb, the highest probability was 34%, after vasopressor initiation, and 25%, after a ratio of saturation to inspired oxygen of <90. The probability after meeting the threshold from randomized trials was 9% (MIMIC) and 13% (AmsterdamUMCdb). In MIMIC, a race/ethnicity of Black (OR, 0.75; 95% CrI, 0.57-0.96) or Asian (OR, 0.6; 95% CrI, 0.35-0.95) compared with White was associated with decreased probability of invasive ventilation after meeting a threshold. Conclusions: The probability of invasive ventilation within 3 hours of meeting physiologic thresholds was low and associated with patient race/ethnicity.

Prediction tool Development and Implementation in pharmacy praCTice (PreDICT) proposed guidance

PURPOSE

Proposed guidance is presented for Prediction tool Development and Implementation in pharmacy praCTice (PreDICT). This guidance aims to assist pharmacists and their collaborators with planning, developing, and implementing custom risk prediction tools for use by pharmacists in their own health systems or practice settings. We aimed to describe general considerations that would be relevant to most prediction tools designed for use in health systems or other pharmacy practice settings.

SUMMARY

The PreDICT proposed guidance is organized into 3 sequential phases: (1) planning, (2) development and validation, and (3) testing and refining prediction tools for real-world use. Each phase is accompanied by a checklist of considerations designed to be used by pharmacists or their trainees (eg, residents) during the planning or conduct of a prediction tool project. Commentary and a worked example are also provided to highlight some of the most relevant and impactful considerations for each phase.

CONCLUSION

The proposed guidance for PreDICT is a pharmacist-focused set of checklists for planning, developing, and implementing prediction tools in pharmacy practice. The list of considerations and accompanying commentary can be used as a reference by pharmacists or their trainees before or during the completion of a prediction tool project.

COVID-19 as a risk factor for long-term mortality in patients managed by the emergency medical system: A prospective, multicenter, ambulance-based cohort study

Introduction

COVID-19 has initially been studied in terms of an acute-phase disease, although recently more attention has been given to the long-term consequences. In this study, we examined COVID-19 as an independent risk factor for long-term mortality in patients with acute illness treated by EMS (emergency medical services) who have previously had the disease against those who have not had the disease.

Methods

A prospective, multicenter, ambulance-based, ongoing study was performed with adult patients with acute disease managed by EMS and transferred with high priority to the emergency department (ED) as study subjects. The study involved six advanced life support units, 38 basic life support units, and five emergency departments from Spain. Sociodemographic inputs, baseline vital signs, pre-hospital blood tests, and comorbidities, including COVID-19, were collected. The main outcome was long-term mortality, which was classified into 1-year all-cause mortality and 1-year in- and out-of-hospital mortality. To compare both the patients with COVID-19 vs. patients without COVID-19 and to compare survival vs non-survival, two main statistical analyses were performed, namely, a longitudinal analysis (Cox regression) and a logistic regression analysis.

Results

Between 12 March 2020 and 30 September 2021, a total of 3,107 patients were included in the study, with 2,594 patients without COVID-19 and 513 patients previously suffering from COVID-19. The mortality rate was higher in patients with COVID-19 than in patients without COVID-19 (31.8 vs. 17.9%). A logistic regression showed that patients previously diagnosed with COVID-19 presented higher rates of nursing home residency, a higher number of breaths per minute, and suffering from connective disease, dementia, and congestive heart failure. The longitudinal analysis showed that COVID-19 was a risk factor for mortality [hazard ratio 1.33 (1.10-1.61); p < 0.001].

Conclusion

The COVID-19 group presented an almost double mortality rate compared with the non-COVID-19 group. The final model adjusted for confusion factors suggested that COVID-19 was a risk factor for long-term mortality.

Canadian C-spine Rule versus NEXUS in Screening of Clinically Important Traumatic Cervical Spine Injuries; a systematic review and meta-analysis

Introduction

The Canadian C-spine Rule (CCR) and the National Emergency X-Radiography Utilization Study (NEXUS) are two criteria designed to rule-out clinically important traumatic cervical Spinal Cord Injury (SCI). In this systematic review and meta-analysis, we reviewed the articles comparing the performance of these two models.

Methods

Search was done in Medline, Embase, Scopus and Web of Science until June 2022. Observational studies with direct comparison of CCR and NEXUS criteria in detection of clinically important cervical SCI were included. Two independent reviewers screened the relevant articles and summarized the data. Certainty of evidence was assessed based on QUADAS-2. Data were recorded as true positive, true negative, false positive, and false negative. Then, using "diagma" package and applying weighted random effect model, area under the receiver operating characteristic (ROC) curve (AUC), sensitivity, specificity, negative likelihood ratio, positive likelihood ratio, and diagnostic odds ratio (DOR) were calculated with 95% confidence interval (95% CI).

Results

We included 5 studies with direct comparison. Area under the ROC curve of NEXUS in screening of patients with clinically important cervical SCI was 0.708 (95% CI: 0.647 to 0.762). Pooled sensitivity and specificity of NEXUS criteria in screening of patients with clinically important cervical SCI were 0.899 (95% CI: 0.845 to 0.936) and 0.398 (95% CI: 0.315 to 0.488). The positive and negative likelihood ratios of NEXUS were 1.494 (95% CI: 1.146 to 1.949) and 0.254 (95% CI: 1.155 to 1.414), respectively. Diagnostic odds ratio of NEXUS was 5.894 (95% CI: 3.372 to 10.305). Furthermore, area under the ROC curve of CCR in screening of clinically important cervical SCI was 0.793 (95% CI: 0.657 to 0.884). Meta-analysis results showed that pooled sensitivity of CCR criteria in screening of patients with clinically important cervical SCI was 0.987 (95% CI: 0.957 to 0.996) and specificity was 0.167 (95% CI: 0.073 to 0.336). The positive and negative likelihood ratios of CCR were 1.184 (95% CI: 0.837 to 1.675) and 0.081 (95% CI: 0.021 to 0.308), respectively. Diagnostic odds ratio of CCR was 14.647 (95% CI: 3.678 to 58.336).

Conclusion

Based on studies, both CCR and NEXUS were sensitive rules that have the potential to reduce unnecessary imaging in cervical spine trauma patients. However, the low specificity and false-positive results of both of these tools indicate that many people will continue to undergo unnecessary imaging after screening of cervical SCI using these tools. In this meta-analysis, CCR appeared to have better screening accuracy.

Application of the pulmonary embolism rule-out criteria (PERC rule) and age-adjusted D-Dimer in patients undergoing computed tomography pulmonary angiography for diagnosis of pulmonary embolism

Background

Diagnosis of pulmonary embolism (PE) constitutes a challenge for practitioners. Current practice involves use of pre-test probability prediction rules. Several strategies to optimize this process have been explored.

Objectives

To explore whether application of the pulmonary embolism rule-out criteria (PERC rule) and age-adjusted D-dimer (DD) would have reduced the number of computed tomography pulmonary angiography (CTPA) examinations performed in patients with suspected PE.

Methods

A retrospective cross-sectional study of adult patients taken for CTPA under suspicion of PE in 2018 and 2020. The PERC rule and age-adjusted DD were applied. The number of cases without indications for imaging studies was estimated and the operational characteristics for diagnosis of PE were calculated.

Results

302 patients were included. PE was diagnosed in 29.8%. Only 27.2% of 'not probable' cases according to the Wells criteria had D-dimer assays. Age adjustment would have reduced tomography use by 11.1%, with an AUC of 0.5. The PERC rule would have reduced use by 7%, with an AUC of 0.72.

Conclusions

Application of age-adjusted D-dimer and the PERC rule to patients taken for CTPA because of suspected PE seems to reduce the number of indications for the procedure.

Translation, adaptation, and validation of the MeNTS score in colombian population: The MeNTS Col score

Objective

To evaluate and validate the medically necessary and time sensitive score by testing the variables, in order to create a surgical preoperative score for procedure prioritization in COVID-19 pandemic in Colombia.

Methods

A multicenter retrospective cross-sectional study of instrument validation with a cultural adaptation and translation into the Spanish language was carried out in Bogota, Colombia. Patients over 18 years of age who had undergone elective procedures of general surgery and subspecialties were included. The translation of the medically necessary and time sensitive score into Spanish was performed independently by two bilingual surgeons fluent in both English and Spanish. A final version of the Spanish questionnaire (MeNTS Col) for testing was then produced by an expert committee. After translation and cultural adaptation, it was submitted to evaluate the psychometric properties of the medically necessary and time sensitive score. Cronbach's α was used to represent and evaluate the internal consistency and assess reliability.

Results

A total of 172 patients were included, with a median age of 54 years; of which 96 (55.8%) patients were females. The vast majority of patients were treated for general surgery (n = 60) and colon and rectal surgery (n = 31). The evaluation of the internal consistency of the scale items in Spanish version was measured, and values of 0.5 for 0.8 were obtained. In the reliability and validation process, Cronbach's α values in all items remained higher than 0.7. The new MeNTS Col model was analyzed, and a result of 0.91 was obtained.

Conclusions

The Spanish version of the medically necessary and time sensitive, the MeNTS Col score, and its respective Spanish translation perform similarly to the original version. Therefore, they can be useful and reproducible in Latin American countries.

Current Diagnosis and Management of Acute Pulmonary Embolism: A Strategy for General Practitioners in Emergency Department

Pulmonary embolism (PE) is a disease with a relatively good prognosis when diagnosed and treated properly. This review aims to analyse available data and combine them into algorithms that physicians can use in the emergency department for quick decision-making in diagnosing and treating PE. The available data show that PE can be excluded through highly sensitive clinical decision rules, i.e. Pulmonary Embolism Rule-Out Criteria (PERC), Wells criteria, and Revised Geneva criteria, combined with D-dimer assessment. In cases where PE could not be excluded through the mentioned strategies, imaging modalities, such as compression ultrasonography (CUS), computed tomographic pulmonary angiography (CTPA), and planar ventilation/perfusion (V/Q) scan, are indicated for a definite diagnosis. Once a diagnosis has been made, treatment of PE depends on its mortality risk as patients are divided into low-, intermediate-, and high-risk cases. High-risk cases are treated for their hemodynamic instability, given parenteral or oral anticoagulant therapy, and are indicated for reperfusion therapy. Intermediate-risk PE is only given parenteral or oral anticoagulants and reperfusion is indicated when anticoagulants fail. Low-risk cases are given oral anticoagulants and based on the Hestia criteria, patients may be discharged and treated as outpatients.

Bacteremia Prediction With Prognostic Scores and a Causal Probabilistic Network - A Cohort Study of Emergency Department Patients

BACKGROUND

Physicians tend to overestimate patients' pretest probability of having bacteremia. The low yield of blood cultures and contaminants is associated with significant financial cost, as well as increased length of stay and unnecessary antibiotic treatment.

OBJECTIVE

This study examined the abilities of the National Early Warning Score (NEWS), the Quick Sequential Organ Failure Assessment (qSOFA), the Modified Sequential Organ Failure Assessment (mSOFA), and two versions of the causal probabilistic network, SepsisFinder™ (SF) to predict bacteremia in adult emergency department (ED) patients.

METHODS

This cohort study included adult ED patients from a large urban, academic tertiary hospital, with blood cultures obtained within 24 h of admission between 2016 and 2017. The outcome measure was true bacteremia. NEWS, qSOFA, mSOFA, and the two versions of SF score were calculated for all patients based on the first available full set of vital signs within 2 h and laboratory values within 6 h after drawing the blood cultures. Area under the receiver operating characteristic curve (AUROC) was calculated for each scoring system.

RESULTS

The study included 3106 ED patients, of which 199 (6.4%) patients had true bacteremia. The AUROCs for prediction of bacteremia were: NEWS = 0.65, qSOFA = 0.60, SF I = 0.65, mSOFA = 0.71, and SF II = 0.80.

CONCLUSIONS

Scoring systems using only vital signs, NEWS, and SF I showed moderate abilities in predicting bacteremia, whereas qSOFA performed poorly. Scoring systems using both vital signs and laboratory values, mSOFA and especially SF II, showed good abilities in predicting bacteremia.

Primer on Logistic Regression for Emergency Care Researchers

BACKGROUND

Logistic regression plays a fundamental role in the production of decision rules, risk assessment, and in establishing cause and effect relationships. This primer is aimed at novice researchers with minimal statistical expertise.

OBJECTIVE

Introduce the logit equation and provide a hands-on example to facilitate understanding of its benefits and limitations.

DISCUSSION

This primer reviews the mathematical basis of a logit equation by comparing and contrasting it with the simple straight-line (linear) equation. After gaining an understanding of the meaning of beta coefficients, readers are encouraged to download a free statistical program and database to produce a logistic regression analysis. Using this example, the narrative then discusses commonly used methods to describe model fitness, including the C-statistic, chi square, Akaike and Bayesian Information Criteria, McFadden's pseudo R², and the Hosmer-Lemeshow test. The authors provide a how-to discussion for variable selection and estimate of sample size. However, logistic regression alone can seldom establish causal inference without further steps to explore the often complex relationship amongst variables and outcomes, such as with the use of a directed acyclic graphs. We present key elements that generally should be considered when appraising an article that uses logistic regression. This primer provides a basic understanding of the theory, hands-on construction, model analysis, and limitations of logistic regression in emergency care research.

CONCLUSIONS

Logistic regression can provide information about the association of independent variables with important clinical outcomes, which can be the first step to show predictiveness or causation of variables on the outcomes of interest. © 2022 Elsevier Inc.

[Machine Learning vs. Statistical Model for Prediction Modelling: Application in Medical Imaging Research]

Clinical prediction models has been increasingly published in radiology research. In particular, as a radiomics research is being actively conducted, the prediction model is developed based on the traditional statistical model, as well as machine learning, to account for the high-dimensional data. In this review, we investigated the statistical and machine learning methods used in clinical prediction model research, and briefly summarized each analytical method for statistical model, machine learning, and statistical learning. Finally, we discussed several considerations for choosing the prediction modeling method.

A Retrospective Analysis of the De Ritis Ratio in Muscle Invasive Bladder Cancer, with Focus on Tumor Response and Long-Term Survival in Patients Receiving Neoadjuvant Chemotherapy and in Chemo Naïve Cystectomy Patients-A Study of a Clinical Multicentre Database

Background: A high pre-treatment De Ritis ratio, the aspartate transaminase/alanine aminotransferase ratio, has been suggested to be of prognostic value for mortality in muscle-invasive bladder cancer (MIBC). Our purpose was to evaluate if a high ratio was associated with mortality and downstaging. Methods: A total of 347 Swedish patients with clinically staged T2-T4aN0M0, with administered neoadjuvant chemotherapy (NAC) or eligible for NAC and undergoing radical cystectomy (RC) 2009−2021, were retrospectively evaluated with a low ratio < 1.3 vs. high ratio > 1.3, by Log Rank test, Cox regression and Mann−Whitney U-test (MWU), SPSS 27. Results: Patients with a high ratio had a decrease of up to 3 years in disease-free survival (DFS), cancer-specific survival (CSS) and overall survival (OS) (p = 0.009, p = 0.004 and p = 0.009) and 5 years in CSS and OS (p = 0.019 and p = 0.046). A high ratio was associated with increased risk of mortality, highest in DFS (HR, 1.909; 95% CI, 1.265−2.880; p = 0.002). No significant relationship between downstaging and a high ratio existed (p = 0.564 MWU). Conclusion: A high pre-treatment De Ritis ratio is on a population level, associated with increased mortality post-RC in endpoints DFS, CSS and OS. Associations decrease over time and require further investigations to determine how strong the associations are as meaningful prognostic markers for long-term mortality in MIBC. The ratio is not suitable for downstaging-prediction.

Regression Model Decreasing the Risk of Femoral Neurovascular Bundle Accidental Puncture

Although most of the adverse events derived from dry needling are minor, avoiding potential hazards for patients including accidental invasion of vessels, ganglia, and nerves is essential to ensure patients’ safety. We aimed to investigate the contribution of predictors explaining the variance of sartorius muscle depth limit at proximal third and middle thigh as these locations lead to an augmented risk of neurovascular bundle invasion during dry needling application. A diagnostic study was conducted on 84 subjects to calculate the accuracy of a prediction model for sartorius depth, as assessed with ultrasound imaging, based on sex, age, height, weight, body mass index (BMI), thigh perimeter, and length. After calculating a correlation matrix, a multiple linear regression analysis was performed to detect those variables contributing to the sartorius deep limit in both locations. Although males showed greater thigh perimeter than women (p < 0.001), the deep limit of the sartorius muscle was significantly more superficial for both the proximal third (p = 0.003) and the mid-third (p = 0.004) points. No side-to-side anthropometric differences were found (p > 0.05). In addition, we found sartorius muscle depth to be associated with the proximal and mid-third girth, gender, height, and BMI (all, p < 0.01). Gender, proximal-third girth, and BMI explained 51.1% and 42.6% of the variance for the sartorius deep limit at the proximal and the mid-third, respectively. This study analyzed whether anthropometric features could predict sartorius muscle depth in healthy participants for assisting clinicians in choosing the optimal needle length to avoid accidental femoral bundle puncture.

Network Analysis for Better Understanding the Complex Psycho-Biological Mechanisms behind Fibromyalgia Syndrome

The aim of this study was to assess potential associations between sensory, cognitive, health-related, and physical variables in women with fibromyalgia syndrome (FMS) using a network analysis for better understanding the complexity of psycho-biological mechanisms. Demographic, clinical, pressure pain threshold (PPT), health-related, physical, and psychological/cognitive variables were collected in 126 women with FMS. A network analysis was conducted to quantify the adjusted correlations between the modeled variables and to assess the centrality indices (i.e., the degree of connection with other symptoms in the network and the importance in the system modeled as a network. This model showed several local associations between the variables. Multiple positive correlations between PPTs were observed, being the strongest weight between PPTs over the knee and tibialis anterior (ρ: 0.28). Catastrophism was associated with higher hypervigilance (ρ: 0.23) and lower health-related EuroQol-5D (ρ: −0.24). The most central variables were PPT over the tibialis anterior (the highest strength centrality), hand grip (the highest harmonic centrality) and Time Up and Go (the highest betweenness centrality). This study, applying network analysis to understand the complex mechanisms of women with FMS, supports a model where sensory-related, psychological/cognitive, health-related, and physical variables are connected. Implications of the current findings, e.g., developing treatments targeting these mechanisms, are discussed.

The Ability of Clinical Decision Rules to Detect Peripheral Arterial Disease: A Narrative Review

Peripheral arterial disease (PAD) is a common cause of lower extremity wound. Consequently, PAD leads to a cause of leg amputation nowadays, especially in diabetic patients. In general practice (GP), confrontation with PAD prevention is a challenge. In general, ankle-brachial index (ABI) measurement can be used as a PAD diagnostic tool, but this takes some time. The tool is not generally available and this need to train healthcare workers to perform. Multiple independent predictors developed the diagnostic prediction model known as clinical decision rules (CDRs) to identify patients with high-risk PAD. This might therefore limit the number of patients (only high-risk patients) to refer for ABI evaluation. This narrative review summarized existing CDRs for PAD.

Clinical prediction of laboratory-confirmed influenza in adults with influenza-like illness in primary care. A randomized controlled trial secondary analysis in 15 European countries

BACKGROUND

Clinical findings do not accurately predict laboratory diagnosis of influenza. Early identification of influenza is considered useful for proper management decisions in primary care.

OBJECTIVE

We evaluated the diagnostic value of the presence and the severity of symptoms for the diagnosis of laboratory-confirmed influenza infection among adults presenting with influenza-like illness (ILI) in primary care.

METHODS

Secondary analysis of patients with ILI who participated in a clinical trial from 2015 to 2018 in 15 European countries. Patients rated signs and symptoms as absent, minor, moderate, or major problem. A nasopharyngeal swab was taken for microbiological identification of influenza and other microorganisms. Models were generated considering (i) the presence of individual symptoms and (ii) the severity rating of symptoms.

RESULTS

A total of 2,639 patients aged 18 or older were included in the analysis. The mean age was 41.8 ± 14.7 years, and 1,099 were men (42.1%). Influenza was microbiologically confirmed in 1,337 patients (51.1%). The area under the curve (AUC) of the model for the presence of any of seven symptoms for detecting influenza was 0.66 (95% confidence interval [CI]: 0.65-0.68), whereas the AUC of the symptom severity model, which included eight variables-cough, fever, muscle aches, sweating and/or chills, moderate to severe overall disease, age, abdominal pain, and sore throat-was 0.70 (95% CI: 0.69-0.72).

CONCLUSION

Clinical prediction of microbiologically confirmed influenza in adults with ILI is slightly more accurate when based on patient reported symptom severity than when based on the presence or absence of symptoms.

Development and Validation of a Prediction Model for the Cure of Peritoneal Dialysis-Associated Peritonitis: A Multicenter Observational Study

Aim

Peritoneal dialysis (PD)-associated peritonitis (PDAP) is a severe complication of PD. It is an important issue about whether it can be cured. At present, there is no available prediction model for peritonitis cure. Therefore, this study aimed to develop and validate a prediction model for peritonitis cure in patients with PDAP.

Methods

Patients with PD who developed PDAP from four dialysis centers in Northeast China were followed up. According to the region of PD, data were divided into training and validation datasets. Initially, a nomogram for peritonitis cure was established based on the training dataset. Later, the nomogram performance was assessed by discrimination (C-statistic), calibration, and decision curves.

Results

Totally, 1,011 episodes of peritonitis were included in the final analysis containing 765 in the training dataset and 246 in the validation dataset. During the follow-up period, peritonitis cure was reported in 615 cases from the training dataset and 198 from the validation dataset. Predictors incorporated in the final nomogram included PD duration, serum albumin, antibiotics prior to admission, white cell count in peritoneal dialysate on day 5 (/μl) ≥ 100/μl, and type of causative organisms. The C-statistic values were 0.756 (95% CI: 0.713–0.799) in the training dataset and 0.756 (95% CI: 0.681–0.831) in the validation dataset. The nomogram exhibited favorable performance in terms of calibration in both the training and validation datasets.

Conclusion

This study develops a practical and convenient nomogram for the prediction of peritonitis cure in patients with PDAP, which assists in clinical decision-making.

CURB-65, qSOFA, and SIRS Criteria in Predicting In-Hospital Mortality of Critically Ill COVID-19 Patients; a Prognostic Accuracy Study

INTRODUCTION

Outcome prediction of intensive care unit (ICU)-admitted patients is one of the important issues for physicians. This study aimed to compare the accuracy of Quick Sequential Organ Failure Assessment (qSOFA), Confusion, Urea, Respiratory Rate, Blood Pressure and Age Above or Below 65 Years (CURB-65), and Systemic Inflammatory Response Syndrome (SIRS) scores in predicting the in-hospital mortality of COVID-19 patients.

METHODS

This prognostic accuracy study was performed on 225 ICU-admitted patients with a definitive diagnosis of COVID-19 from July to December 2021 in Tehran, Iran. The patients' clinical characteristics were evaluated at the time of ICU admission, and they were followed up until discharge from ICU. The screening performance characteristics of CURB-65, qSOFA, and SIRS in predicting their mortality was compared.

RESULTS

225 patients with the mean age of 63.27±14.89 years were studied (56.89% male). The in-hospital mortality rate of this series of patients was 39.10%. The area under the curve (AUC) of SIRS, CURB-65, and qSOFA were 0.62 (95% CI: 0.55 - 0.69), 0.66 (95% CI: 0.59 - 0.73), and 0.61(95% CI: 0.54 - 0.67), respectively (p = 0.508). In cut-off ≥1, the estimated sensitivity values of SIRS, CURB-65, and qSOFA were 85.23%, 96.59%, and 78.41%, respectively. The estimated specificity of scores were 34.31%, 6.57%, and 38.69%, respectively. In cut-off ≥2, the sensitivity values of SIRS, CURB-65, and qSOFA were evaluated as 39.77%, 87.50%, and 15.91%, respectively. Meanwhile, the specificity of scores were 72.99%, 34.31%, and 92.70%.

CONCLUSIONS

It seems that the performance of SIRS, CURB-65, and qSOFA is similar in predicting the ICU mortality of COVID-19 patients. However, the sensitivity of CURB-65 is higher than qSOFA and SIRS.

Factors Associated with 24-Hour Clinical Outcome of Emergency Patients; a Cohort Study

Introduction

Pre-hospital and in-hospital emergency management play an important role in quality of care for emergency patients. This prospective cohort study aimed to determine factors associated with the 24-hour clinical outcome of emergency patients.

Methods

The sample included 1,630 patients, randomly selected through multi-stage stratified sampling from 13 hospitals in 13 provinces of Thailand. Data were collected during January-November 2018. Clinical outcome was determined using pre-arrest sign score. Data were analyzed via ordinal multivariate regression analysis.

Results

Factors influencing 24-hour clinical outcome of emergency patients were age (OR: 0.965; 95% CI: 0.96-0.97), having coronary vascular disease (CAD) (OR: 1.41; 95% CI: 1.05-1.88), and severity of illness based on Rapid Emergency Medical Score (REMS) (OR:1.09; 95% CI: 1.05-1.15). Self-transportation and being transported by emergency medical service ambulance with non-advanced life support (EMS-Non-ALS) did not influence clinical outcome when compared to EMS-ALS transport. Being transported from a community hospital increased pre-arrest sign score 1.78 times when compared to EMS-ALS (OR: 1.78; 95% CI: 1.17-2.72). Increased transportation distance increased the risk of poor clinical outcome (OR: 1.01; 95% CI: 1.002-1.011). Length of stay in emergency department (ED-LOS) more than 4 hours (OR: 0.21; 95% CI: 0.15-0.29) and between 2-4 hours (OR: 0.60; 95% CI: 0.47-0.75) decreased the risk of poor clinical outcome when compared to ED-LOS less than 2 hours.

Conclusion

Having CAD, severity of illness, increased transport distance, and ED-LOS less than 2 hours were found to negatively influence 24-hour clinical outcome of emergency patients.

Development and Validation of Models to Predict Pathological Outcomes of Radical Prostatectomy in Regional and National Cohorts

PURPOSE

Prediction models are recommended by national guidelines to support clinical decision making in prostate cancer. Existing models to predict pathological outcomes of radical prostatectomy (RP)-the Memorial Sloan Kettering (MSK) models, Partin tables, and the Briganti nomogram-have been developed using data from tertiary care centers and may not generalize well to other settings.

MATERIALS AND METHODS

Data from a regional cohort (Michigan Urological Surgery Improvement Collaborative [MUSIC]) were used to develop models to predict extraprostatic extension (EPE), seminal vesicle invasion (SVI), lymph node invasion (LNI), and nonorgan-confined disease (NOCD) in patients undergoing RP. The MUSIC models were compared against the MSK models, Partin tables, and Briganti nomogram (for LNI) using data from a national cohort (Surveillance, Epidemiology, and End Results [SEER] registry).

RESULTS

We identified 7,491 eligible patients in the SEER registry. The MUSIC model had good discrimination (SEER AUC EPE: 0.77; SVI: 0.80; LNI: 0.83; NOCD: 0.77) and was well calibrated. While the MSK models had similar discrimination to the MUSIC models (SEER AUC EPE: 0.76; SVI: 0.80; LNI: 0.84; NOCD: 0.76), they overestimated the risk of EPE, LNI, and NOCD. The Partin tables had inferior discrimination (SEER AUC EPE: 0.67; SVI: 0.76; LNI: 0.69; NOCD: 0.72) as compared to other models. The Briganti LNI nomogram had an AUC of 0.81 in SEER but overestimated the risk.

CONCLUSIONS

New models developed using the MUSIC registry outperformed existing models and should be considered as potential replacements for the prediction of pathological outcomes in prostate cancer.

Prediction Model Using Readily Available Clinical Data for Colorectal Cancer in Chinese Population

BACKGROUND

In China, health screening has become common, although colonoscopy is not always available or acceptable. We sought to develop a prediction model of colorectal cancer (CRC) for health screening population based on readily available clinical data to reduce labor and economic costs.

METHODS

We conducted a cross-sectional study based on a health screening population in Karamay Central Hospital. By collecting clinical data and basic information from participants, we identified independent risk factors and established a prediction model of CRC. Internal and external validation, calibration plot, and decision curve analysis were employed to test discriminating ability, calibration ability, and clinical practicability.

RESULTS

Independent risk factors of CRC, which were readily available in basic public health institutions, included high-density lipoprotein cholesterol, male sex, total cholesterol, advanced age, and hemoglobin. These factors were successfully incorporated into the prediction model (AUC 0.740, 95% CI 0.713-0.767). The model demonstrated a high degree of discrimination and calibration, in addition to a high degree of clinical practicability in high-risk people.

CONCLUSIONS

The prediction model exhibits good discrimination and calibration and is pragmatic for CRC screening in rural areas and basic public health institutions.

Syncope Prediction Scores in the Emergency Department

Syncope is a common clinical presentation defined as a transient loss of consciousness (TLOC) due to cerebral hypoperfusion, characterized by a rapid onset, short duration, and spontaneous complete recovery. Different clinical decision rules (CDRs) and risk stratification scores have been developed to predict short- and long-term risks for adverse outcomes after syncope. The central theme of these prediction systems is consistent with the ESC syncope guidelines. Initial assessment according to the ESC guideline is essential until an optimal and well-validated risk score is available. The focus should be accurate risk stratification to allow prevention of adverse outcomes and optimize the use of limited healthcare resources. In this review article, we summarize and critically appraise the evidence regarding the CDRs for patients presenting with syncope.

Improving appropriate use of anticoagulants in hospitalised patients: a pharmacist-led Check of Medication Appropriateness intervention

AIM

Inappropriate anticoagulant use increases the risk of bleeding and thrombotic events. We implemented clinical decision rules to promote judicious medication use, as part of the 'Check of Medication Appropriateness' (CMA). The CMA concerns a pharmacist-led review service, targeting potentially inappropriate prescriptions (PIPs). In this analysis, we aimed to evaluate the impact of the CMA on anticoagulant prescribing.

METHODS

The number of anticoagulant-related PIPs was evaluated before and after implementation of the intervention in a quasi-experimental interrupted time series analysis. The pre-implementation cohort received usual care. The anticoagulant-focused CMA, comprising 13 clinical rules pertaining to anticoagulation therapies, was implemented in the post-implementation cohort. Segmented regression analysis was used to assess the impact of the intervention on the number of residual PIPs. A residual PIP was defined as a PIP which persisted up to 48h after the CMA intervention. Total number of recommendations and acceptance rate were documented for the 2-year post-implementation period.

RESULTS

Pre-implementation, we observed 501 PIPs in 466 inpatients on 36 days, with a median proportion of 78.5% (range: 46.2%-100%) residual PIPs per day. Post-implementation, 538 PIPs were detected in 485 patients over the same number of days. The CMA intervention reduced the median proportion to 18.2% (range: 0-100%) per day. The effect coincided with an immediate relative reduction of 70% (95%CI 0.19-0.46) in anticoagulant-related residual PIPs. Post-implementation, 2778 recommendations were provided and 75.1% were accepted.

CONCLUSION

Our CMA approach significantly reduced anticoagulant-related PIPs. Implementing a pharmacist-led intervention, based on clinical rules, may support safer prescribing of anticoagulants.

External validation of the modified LACE+, LACE+, and LACE scores to predict readmission or death after hospital discharge

BACKGROUND

Unplanned hospital readmissions are common adverse events. The LACE+ score has been used to identify patients at the highest risk of unplanned readmission or death, yet the external validity of this score remains uncertain.

METHODS

We constructed a cohort of patients admitted to hospital between 1 October 2014 and 31 January 2017 using population-based data from British Columbia (Canada). The primary outcome was a composite of urgent hospital readmission or death within 30 days of index discharge. The primary analysis sought to optimize clinical utility and international generalizability by focusing on the modified LACE+ (mLACE+) score, a variation of the LACE+ score which excludes the Case Mix Group score. Predictive performance was assessed using model calibration and discrimination.

RESULTS

Among 368,154 hospitalized individuals, 31,961 (8.7%) were urgently readmitted and 5428 (1.5%) died within 30 days of index discharge (crude composite risk of readmission or death, 9.95%). The mLACE+ score exhibited excellent calibration (calibration-in-the-large and calibration slope no different than ideal) and adequate discrimination (c-statistic, 0.681; 95%CI, 0.678 to 0.684). Higher risk dichotomized mLACE+ scores were only modestly associated with the primary outcome (positive likelihood ratio 1.95, 95%CI 1.93 to 1.97). Predictive performance of the mLACE+ score was similar to that of the LACE+ and LACE scores.

CONCLUSION

The mLACE+, LACE+ and LACE scores predict hospital readmission with excellent calibration and adequate discrimination. These scores can be used to target interventions designed to prevent unplanned hospital readmission.

Predicting Prolonged Length of ICU Stay through Machine Learning

This study aimed to construct machine learning (ML) models for predicting prolonged length of stay (pLOS) in intensive care units (ICU) among general ICU patients. A multicenter database called eICU (Collaborative Research Database) was used for model derivation and internal validation, and the Medical Information Mart for Intensive Care (MIMIC) III database was used for external validation. We used four different ML methods (random forest, support vector machine, deep learning, and gradient boosting decision tree (GBDT)) to develop prediction models. The prediction performance of the four models were compared with the customized simplified acute physiology score (SAPS) II. The area under the receiver operation characteristic curve (AUROC), area under the precision-recall curve (AUPRC), estimated calibration index (ECI), and Brier score were used to measure performance. In internal validation, the GBDT model achieved the best overall performance (Brier score, 0.164), discrimination (AUROC, 0.742; AUPRC, 0.537), and calibration (ECI, 8.224). In external validation, the GBDT model also achieved the best overall performance (Brier score, 0.166), discrimination (AUROC, 0.747; AUPRC, 0.536), and calibration (ECI, 8.294). External validation showed that the calibration curve of the GBDT model was an optimal fit, and four ML models outperformed the customized SAPS II model. The GBDT-based pLOS-ICU prediction model had the best prediction performance among the five models on both internal and external datasets. Furthermore, it has the potential to assist ICU physicians to identify patients with pLOS-ICU risk and provide appropriate clinical interventions to improve patient outcomes.

The CoVID-TE risk assessment model for venous thromboembolism in hospitalized patients with cancer and COVID-19

BACKGROUND

Hospitalized patients with COVID-19 have increased risks of venous (VTE) and arterial thromboembolism (ATE). Active cancer diagnosis and treatment are well-known risk factors; however, a risk assessment model (RAM) for VTE in patients with both cancer and COVID-19 is lacking.

OBJECTIVES

To assess the incidence of and risk factors for thrombosis in hospitalized patients with cancer and COVID-19.

METHODS

Among patients with cancer in the COVID-19 and Cancer Consortium registry (CCC19) cohort study, we assessed the incidence of VTE and ATE within 90 days of COVID-19-associated hospitalization. A multivariable logistic regression model specifically for VTE was built using a priori determined clinical risk factors. A simplified RAM was derived and internally validated using bootstrap.

RESULTS

From March 17, 2020 to November 30, 2020, 2804 hospitalized patients were analyzed. The incidence of VTE and ATE was 7.6% and 3.9%, respectively. The incidence of VTE, but not ATE, was higher in patients receiving recent anti-cancer therapy. A simplified RAM for VTE was derived and named CoVID-TE (Cancer subtype high to very-high risk by original Khorana score +1, VTE history +2, ICU admission +2, D-dimer elevation +1, recent systemic anti-cancer Therapy +1, and non-Hispanic Ethnicity +1). The RAM stratified patients into two cohorts (low-risk, 0-2 points, n = 1423 vs. high-risk, 3+ points, n = 1034) where VTE occurred in 4.1% low-risk and 11.3% high-risk patients (c statistic 0.67, 95% confidence interval 0.63-0.71). The RAM performed similarly well in subgroups of patients not on anticoagulant prior to admission and moderately ill patients not requiring direct ICU admission.

CONCLUSIONS

Hospitalized patients with cancer and COVID-19 have elevated thrombotic risks. The CoVID-TE RAM for VTE prediction may help real-time data-driven decisions in this vulnerable population.

Building and evaluating suicide attempt prediction models using risk factors

This study identified risk factors of suicide attempts for the purpose of building prediction models and evaluating their performance. The participants of this secondary data analysis study were 11 671 adults aged 19 years or older. The prediction models consisted of risk factors identified through multiple logistic regression analysis, and performance was analyzed in terms of calibration, discrimination, and clinical usefulness. The risk factors for suicide attempts were suicide plan and suicidal ideation for males, and suicide plan and depression diagnosis for females. The prediction models constructed with these risk factors showed good calibration and discrimination, with over 0.90 of the area under the curves. At the cutoff point of 0.5%, the sensitivity of the full model was 90.9% for males and 82.4% for females. The net benefit was positive at a threshold probability under 30% for males and 40% for females. Given the acceptable performance of the suicide attempt prediction models, they can be used to assess suicide attempt risk and detect the population at high risk in the community at an early stage, with limited human resources.