Search

Objectives: Prediction models may facilitate risk-based management of health care conditions. In a large cluster-randomized trial, presenting calculated risks of postoperative nausea and vomiting (PONV) to physicians (assistive approach) increased risk-based management of PONV. This increase did not improve patient outcome-that is, PONV incidence. This prompted us to explore how prediction tools guide the decision-making process of physicians., Study Design and Setting: Using mixed methods, we interviewed eight physicians to understand how predicted risks were perceived by the physicians and how they influenced decision making. Subsequently, all 57 physicians of the trial were surveyed for how the presented risks influenced their perceptions., Results: Although the prediction tool made physicians more aware of PONV prevention, the physicians reported three barriers to use predicted risks in their decision making. PONV was not considered an outcome of utmost importance; decision making on PONV prophylaxis was mostly intuitive rather than risk based; prediction models do not weigh benefits and risks of prophylactic drugs., Conclusion: Combining probabilistic output of the model with their clinical experience may be difficult for physicians, especially when their decision-making process is mostly intuitive. Adding recommendations to predicted risks (directive approach) was considered an important step to facilitate the uptake of a prediction tool., (Copyright © 2016 Elsevier Inc. All rights reserved.)

Objectives: This study aims to investigate the influence of the amount of clustering [intraclass correlation (ICC) = 0%, 5%, or 20%], the number of events per variable (EPV) or candidate predictor (EPV = 5, 10, 20, or 50), and backward variable selection on the performance of prediction models., Study Design and Setting: Researchers frequently combine data from several centers to develop clinical prediction models. In our simulation study, we developed models from clustered training data using multilevel logistic regression and validated them in external data., Results: The amount of clustering was not meaningfully associated with the models' predictive performance. The median calibration slope of models built in samples with EPV = 5 and strong clustering (ICC = 20%) was 0.71. With EPV = 5 and ICC = 0%, it was 0.72. A higher EPV related to an increased performance: the calibration slope was 0.85 at EPV = 10 and ICC = 20% and 0.96 at EPV = 50 and ICC = 20%. Variable selection sometimes led to a substantial relative bias in the estimated predictor effects (up to 118% at EPV = 5), but this had little influence on the model's performance in our simulations., Conclusion: We recommend at least 10 EPV to fit prediction models in clustered data using logistic regression. Up to 50 EPV may be needed when variable selection is performed., (Copyright © 2015 Elsevier Inc. All rights reserved.)

Objective: To validate all diagnostic prediction models for ruling out pulmonary embolism that are easily applicable in primary care., Design: Systematic review followed by independent external validation study to assess transportability of retrieved models to primary care medicine., Setting: 300 general practices in the Netherlands., Participants: Individual patient dataset of 598 patients with suspected acute pulmonary embolism in primary care., Main Outcome Measures: Discriminative ability of all models retrieved by systematic literature search, assessed by calculation and comparison of C statistics. After stratification into groups with high and low probability of pulmonary embolism according to pre-specified model cut-offs combined with qualitative D-dimer test, sensitivity, specificity, efficiency (overall proportion of patients with low probability of pulmonary embolism), and failure rate (proportion of pulmonary embolism cases in group of patients with low probability) were calculated for all models., Results: Ten published prediction models for the diagnosis of pulmonary embolism were found. Five of these models could be validated in the primary care dataset: the original Wells, modified Wells, simplified Wells, revised Geneva, and simplified revised Geneva models. Discriminative ability was comparable for all models (range of C statistic 0.75-0.80). Sensitivity ranged from 88% (simplified revised Geneva) to 96% (simplified Wells) and specificity from 48% (revised Geneva) to 53% (simplified revised Geneva). Efficiency of all models was between 43% and 48%. Differences were observed between failure rates, especially between the simplified Wells and the simplified revised Geneva models (failure rates 1.2% (95% confidence interval 0.2% to 3.3%) and 3.1% (1.4% to 5.9%), respectively; absolute difference -1.98% (-3.33% to -0.74%)). Irrespective of the diagnostic prediction model used, three patients were incorrectly classified as having low probability of pulmonary embolism; pulmonary embolism was diagnosed only after referral to secondary care., Conclusions: Five diagnostic pulmonary embolism prediction models that are easily applicable in primary care were validated in this setting. Whereas efficiency was comparable for all rules, the Wells rules gave the best performance in terms of lower failure rates., (© Hendriksen et al 2015.)

Prediction models are developed to aid health care providers in estimating the probability that a specific outcome or disease is present (diagnostic prediction models) or will occur in the future (prognostic prediction models), to inform their decision making. Prognostic models here also include models to predict treatment outcomes or responses; in the cancer literature often referred to as predictive models. Clinical prediction models have become abundant. Pathology measurement or results are frequently included as predictors in such prediction models, certainly in the cancer domain. Only when full information on all aspects of a prediction modeling study are clearly reported, risk of bias and potential usefulness of the prediction model can be adequately assessed. Many reviews have illustrated that the quality of reports on the development, validation, and/or adjusting (updating) of prediction models, is very poor. Hence, the Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD) initiative has developed a comprehensive and user-friendly checklist for the reporting of studies on, both diagnostic and prognostic, prediction models. The TRIPOD Statement intends to improve the transparency and completeness of reporting of studies that report solely on development, both development and validation, and solely on the validation (with or without updating) of diagnostic or prognostic, including predictive, models.

Background: General practitioners can safely exclude pulmonary embolism (PE) by using the Wells PE rule combined with D-dimer testing., Objective: To compare the accuracy of a strategy using the Wells rule combined with either a qualitative point-of-care (POC) D-dimer test performed in primary care or a quantitative laboratory-based D-dimer test., Methods: We used data from a prospective cohort study including 598 adults suspected of PE in primary care in the Netherlands. General practitioners scored the Wells rule and carried out a qualitative POC test. All patients were referred to hospital for reference testing. We obtained quantitative D-dimer test results as performed in hospital laboratories. The primary outcome was the prevalence of venous thromboembolism in low-risk patients., Results: Prevalence of PE was 12.2%. POC D-dimer test results were available in 582 patients (97%). Quantitative test results were available in 401 patients (67%). We imputed results in 197 patients. The quantitative test and POC test missed one (0.4%) and four patients (1.5%), respectively, with a negative strategy (Wells ≤ 4 points and D-dimer test negative) (P = 0.20). The POC test could exclude 23 more patients (4%) (P = 0.05). The sensitivity and specificity of the Wells rule combined with a POC test were 94.5% and 51.0% and, combined with a quantitative test, 98.6% and 47.2%, respectively., Conclusions: Combined with the Wells PE rule, both tests are safe to use in excluding PE. The quantitative test seemed to be safer than the POC test, albeit not statistically significant. The specificity of the POC test was higher, resulting in more patients in whom PE could be excluded., (© 2015 International Society on Thrombosis and Haemostasis.)

Objective: We aimed to validate the Oudega diagnostic decision rule-which was developed and validated among younger aged primary care patients-to rule-out deep vein thrombosis (DVT) in frail older outpatients., Methods: In older patients (>60 years, either community dwelling or residing in nursing homes) with clinically suspected DVT, physicians recorded the score on the Oudega rule and d-dimer test. DVT was confirmed with a composite reference standard including ultrasonography examination and 3-month follow-up. The proportion of patients with a very low probability of DVT according to the Oudega rule (efficiency), and the proportion of patients with symptomatic venous thromboembolism during 3 months follow-up within this 'very low risk' group (failure rate) was calculated., Results: DVT occurred in 164 (47%) of the 348 study participants (mean age 81 years, 85% residing in nursing homes). The probability of DVT was very low in 69 patients (Oudega score ≤3 points plus a normal d-dimer test; efficiency 20%) of whom four had non-fatal DVT (failure rate 5.8%; 2.3-14%). With a simple revised version of the Oudega rule for older suspected patients, 43 patients had a low risk of DVT (12% of the total population) of whom only one had DVT (failure rate 2.3%; 0.4-12%)., Conclusions: In older suspected patients, application of the original Oudega rule to exclude DVT resulted in a higher failure rate as compared to previous studies. A revised and simplified Oudega strategy specifically developed for elderly suspected patients resulted in a lower failure rate though at the expense of a lower efficiency., (© The Author 2014. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.)

Prediction models are developed to aid health care providers in estimating the probability or risk that a specific disease or condition is present (diagnostic models) or that a specific event will occur in the future (prognostic models), to inform their decision making. However, the overwhelming evidence shows that the quality of reporting of prediction model studies is poor. Only with full and clear reporting of information on all aspects of a prediction model can risk of bias and potential usefulness of prediction models be adequately assessed. The Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis (TRIPOD) Initiative developed a set of recommendations for the reporting of studies developing, validating, or updating a prediction model, whether for diagnostic or prognostic purposes. This article describes how the TRIPOD Statement was developed. An extensive list of items based on a review of the literature was created, which was reduced after a Web-based survey and revised during a 3-day meeting in June 2011 with methodologists, health care professionals, and journal editors. The list was refined during several meetings of the steering group and in e-mail discussions with the wider group of TRIPOD contributors. The resulting TRIPOD Statement is a checklist of 22 items, deemed essential for transparent reporting of a prediction model study. The TRIPOD Statement aims to improve the transparency of the reporting of a prediction model study regardless of the study methods used. The TRIPOD Statement is best used in conjunction with the TRIPOD explanation and elaboration document. To aid the editorial process and readers of prediction model studies, it is recommended that authors include a completed checklist in their submission (also available at www.tripod-statement.org)., (© 2015 Royal College of Obstetricians and Gynaecologists.)

Background: Prediction models are developed to aid healthcare providers in estimating the probability or risk that a specific disease or condition is present (diagnostic models) or that a specific event will occur in the future (prognostic models), to inform their decision-making. However, the overwhelming evidence shows that the quality of reporting of prediction model studies is poor. Only with full and clear reporting of information on all aspects of a prediction model can risk of bias and potential usefulness of prediction models be adequately assessed., Materials and Methods: The Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD) initiative developed a set of recommendations for the reporting of studies developing, validating or updating a prediction model, whether for diagnostic or prognostic purposes. This article describes how the TRIPOD Statement was developed. An extensive list of items based on a review of the literature was created, which was reduced after a Web-based survey and revised during a 3-day meeting in June 2011 with methodologists, healthcare professionals and journal editors. The list was refined during several meetings of the steering group and in e-mail discussions with the wider group of TRIPOD contributors., Results: The resulting TRIPOD Statement is a checklist of 22 items, deemed essential for transparent reporting of a prediction model study. The TRIPOD Statement aims to improve the transparency of the reporting of a prediction model study regardless of the study methods used. The TRIPOD Statement is best used in conjunction with the TRIPOD explanation and elaboration document., Conclusions: To aid the editorial process and readers of prediction model studies, it is recommended that authors include a completed checklist in their submission (also available at www.tripod-statement.org)., (© 2015 Stichting European Society for Clinical Investigation Journal Foundation.)

Prediction models are developed to aid health care providers in estimating the probability or risk that a specific disease or condition is present (diagnostic models) or that a specific event will occur in the future (prognostic models), to inform their decision making. However, the overwhelming evidence shows that the quality of reporting of prediction model studies is poor. Only with full and clear reporting of information on all aspects of a prediction model can risk of bias and potential usefulness of prediction models be adequately assessed. The Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD) Initiative developed a set of recommendations for the reporting of studies developing, validating, or updating a prediction model, whether for diagnostic or prognostic purposes. This article describes how the TRIPOD Statement was developed. An extensive list of items based on a review of the literature was created, which was reduced after a Web based survey and revised during a three day meeting in June 2011 with methodologists, health care professionals, and journal editors. The list was refined during several meetings of the steering group and in e-mail discussions with the wider group of TRIPOD contributors. The resulting TRIPOD Statement is a checklist of 22 items, deemed essential for transparent reporting of a prediction model study. The TRIPOD Statement aims to improve the transparency of the reporting of a prediction model study regardless of the study methods used. The TRIPOD Statement is best used in conjunction with the TRIPOD explanation and elaboration document. To aid the editorial process and readers of prediction model studies, it is recommended that authors include a completed checklist in their submission (also available at www.tripod-statement.org).To encourage dissemination of the TRIPOD Statement, this article is freely accessible on the Annals of Internal Medicine Web site (www.annals.org) and will be also published in BJOG, British Journal of Cancer, British Journal of Surgery, BMC Medicine, The BMJ, Circulation, Diabetic Medicine, European Journal of Clinical Investigation, European Urology, and Journal of Clinical Epidemiology. The authors jointly hold the copyright of this article. An accompanying explanation and elaboration article is freely available only on www.annals.org; Annals of Internal Medicine holds copyright for that article., (© BMJ Publishing Group Ltd 2014.)

Prediction models are developed to aid health care providers in estimating the probability or risk that a specific disease or condition is present (diagnostic models) or that a specific event will occur in the future (prognostic models), to inform their decision making. However, the overwhelming evidence shows that the quality of reporting of prediction model studies is poor. Only with full and clear reporting of information on all aspects of a prediction model can risk of bias and potential usefulness of prediction models be adequately assessed. The Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD) Initiative developed a set of recommendations for the reporting of studies developing, validating, or updating a prediction model, whether for diagnostic or prognostic purposes. This article describes how the TRIPOD Statement was developed. An extensive list of items based on a review of the literature was created, which was reduced after a Web-based survey and revised during a 3-day meeting in June 2011 with methodologists, health care professionals, and journal editors. The list was refined during several meetings of the steering group and in e-mail discussions with the wider group of TRIPOD contributors. The resulting TRIPOD Statement is a checklist of 22 items, deemed essential for transparent reporting of a prediction model study. The TRIPOD Statement aims to improve the transparency of the reporting of a prediction model study regardless of the study methods used. The TRIPOD Statement is best used in conjunction with the TRIPOD explanation and elaboration document. To aid the editorial process and readers of prediction model studies, it is recommended that authors include a completed checklist in their submission (also available at www.tripod-statement.org).

The TRIPOD (Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis) Statement includes a 22-item checklist, which aims to improve the reporting of studies developing, validating, or updating a prediction model, whether for diagnostic or prognostic purposes. The TRIPOD Statement aims to improve the transparency of the reporting of a prediction model study regardless of the study methods used. This explanation and elaboration document describes the rationale; clarifies the meaning of each item; and discusses why transparent reporting is important, with a view to assessing risk of bias and clinical usefulness of the prediction model. Each checklist item of the TRIPOD Statement is explained in detail and accompanied by published examples of good reporting. The document also provides a valuable reference of issues to consider when designing, conducting, and analyzing prediction model studies. To aid the editorial process and help peer reviewers and, ultimately, readers and systematic reviewers of prediction model studies, it is recommended that authors include a completed checklist in their submission. The TRIPOD checklist can also be downloaded from www.tripod-statement.org.

Objectives: To determine whether the Wells clinical prediction rule for pulmonary embolism (PE), which produces a point score based on clinical features and the likelihood of diagnoses other than PE, combined with normal D-dimer testing can be used to exclude PE in older unhospitalized adults., Design: Prospective cohort study., Setting: Primary care and nursing homes., Participants: Older adults (≥60) clinically suspected of having a PE (N = 294, mean age 76, 44% residing in a nursing home)., Measurements: The presence of PE was confirmed using a composite reference standard including computed tomography and 3-month follow-up. The proportion of individuals with an unlikely risk of PE was calculated according to the Wells rule (≤4 points) plus a normal qualitative point-of-care D-dimer test (efficiency) and the presence of symptomatic PE during 3 months of follow-up within these patients (failure rate)., Results: Pulmonary embolism occurred in 83 participants (28%). Eighty-five participants had an unlikely risk according to the Wells rule and a normal D-dimer test (efficiency 29%), five of whom experienced a nonfatal PE during 3 months of follow-up (failure rate = 5.9%, 95% confidence interval (CI) = 2.5-13%). According to a refitted diagnostic strategy for older adults, 69 had a low risk of PE (24%), two of whom had PE (failure rate = 2.9%, 95% CI = 0.8-10%)., Conclusion: The use of the well-known and widely used Wells rule (original or refitted) does not guarantee safe exclusion of PE in older unhospitalized adults with suspected PE. This may lead to discussion among professionals as to whether the original or revised Wells rule is useful for elderly outpatients., (© 2014, Copyright the Authors Journal compilation © 2014, The American Geriatrics Society.)

An important question in the evaluation of an additional risk prediction marker is how to interpret a small increase in the area under the receiver operating characteristic curve (AUC). Many researchers believe that a change in AUC is a poor metric because it increases only slightly with the addition of a marker with a large odds ratio. Because it is not possible on purely statistical grounds to choose between the odds ratio and AUC, we invoke decision analysis, which incorporates costs and benefits. For example, a timely estimate of the risk of later non-elective operative delivery can help a woman in labor decide if she wants an early elective cesarean section to avoid greater complications from possible later non-elective operative delivery. A basic risk prediction model for later non-elective operative delivery involves only antepartum markers. Because adding intrapartum markers to this risk prediction model increases AUC by 0.02, we questioned whether this small improvement is worthwhile. A key decision-analytic quantity is the risk threshold, here the risk of later non-elective operative delivery at which a patient would be indifferent between an early elective cesarean section and usual care. For a range of risk thresholds, we found that an increase in the net benefit of risk prediction requires collecting intrapartum marker data on 68 to 124 women for every correct prediction of later non-elective operative delivery. Because data collection is non-invasive, this test tradeoff of 68 to 124 is clinically acceptable, indicating the value of adding intrapartum markers to the risk prediction model., (Copyright © 2014 John Wiley & Sons, Ltd.)

Background: Increased zinc protoporphyrin (ZPP) levels can indicate iron deficiency and may be predictive for low hemoglobin (Hb) deferral in blood donors. Prediction models for Hb deferral in whole blood donors have already been developed. In this study, we examined if addition of ZPP to these prediction models improves risk estimation of Hb deferral., Study Design and Methods: This study included 4598 Dutch whole blood donors. Information on ZPP levels measured at the previous visit was added to the existing prediction models to estimate the risk of Hb deferral. Models were compared using the following measures: concordance (c)-statistic, continuous net reclassification improvement (NRI), and clinical net benefit (NB)., Results: Seventy-six male donors (2.9%) and 69 female donors (3.5%) were deferred because of a low Hb level. Previous ZPP level was associated with risk of Hb deferral (odds ratio for interquartile range of previous ZPP level, men 2.0 [95% confidence interval {CI}, 1.7-2.3]; women 2.2 [95% CI, 1.9-2.4]) in a multivariable risk model. Addition of ZPP into the models resulted in an increase of the c-statistic from 0.93 to 0.94 for men and from 0.80 to 0.85 for women. The added value of ZPP was confirmed by measures of clinical usefulness. NRI for men was 0.42, and for women, 0.56. At relevant threshold probabilities between 10 and 15%, NB was higher for models considering ZPP., Conclusion: This study shows that ZPP measurements obtained at the previous visit may have added value in the risk prediction of Hb deferral in whole blood donors., (© 2012 American Association of Blood Banks.)

Surveillance of healthcare-associated infections is labor intensive and complex. Discharge coding is an accessible source of information that may support detection of cases. For drain-related meningitis, however, discharge coding data had low sensitivity (32%) and positive predictive value (35%) and could neither replace nor improve existing complex surveillance systems.

Prognostic factor research aims to identify factors associated with subsequent clinical outcome in people with a particular disease or health condition. In this article, the second in the PROGRESS series, the authors discuss the role of prognostic factors in current clinical practice, randomised trials, and developing new interventions, and explain why and how prognostic factor research should be improved.

Purpose: Prediction models for exacerbations in patients with chronic obstructive pulmonary disease (COPD) are scarce. Our aim was to develop and validate a new model to predict exacerbations in patients with COPD., Patients and Methods: The derivation cohort consisted of patients aged 65 years or over, with a COPD diagnosis, who were followed up over 24 months. The external validation cohort consisted of another cohort of COPD patients, aged 50 years or over. Exacerbations of COPD were defined as symptomatic deterioration requiring pulsed oral steroid use or hospitalization. Logistic regression analysis including backward selection and shrinkage were used to develop the final model and to adjust for overfitting. The adjusted regression coefficients were applied in the validation cohort to assess calibration of the predictions and calculate changes in discrimination applying C-statistics., Results: The derivation and validation cohort consisted of 240 and 793 patients with COPD, of whom 29% and 28%, respectively, experienced an exacerbation during follow-up. The final model included four easily assessable variables: exacerbations in the previous year, pack years of smoking, level of obstruction, and history of vascular disease, with a C-statistic of 0.75 (95% confidence interval [CI]: 0.69-0.82). Predictions were well calibrated in the validation cohort, with a small loss in discrimination potential (C-statistic 0.66 [95% CI 0.61-0.71])., Conclusion: Our newly developed prediction model can help clinicians to predict the risk of future exacerbations in individual patients with COPD, including those with mild disease.

Prognostic models are abundant in the medical literature yet their use in practice seems limited. In this article, the third in the PROGRESS series, the authors review how such models are developed and validated, and then address how prognostic models are assessed for their impact on practice and patient outcomes, illustrating these ideas with examples.

Prediction models are increasingly used to complement clinical reasoning and decision making in modern medicine in general, and in the cardiovascular domain in particular. Developed models first and foremost need to provide accurate and (internally and externally) validated estimates of probabilities of specific health conditions or outcomes in targeted patients. The adoption of such models must guide physician's decision making and an individual's behaviour, and consequently improve individual outcomes and the cost-effectiveness of care. In a series of two articles we review the consecutive steps generally advocated for risk prediction model research. This first article focuses on the different aspects of model development studies, from design to reporting, how to estimate a model's predictive performance and the potential optimism in these estimates using internal validation techniques, and how to quantify the added or incremental value of new predictors or biomarkers (of whatever type) to existing predictors. Each step is illustrated with empirical examples from the cardiovascular field.

Background: Clinical probability assessment is combined with d-dimer testing to exclude pulmonary embolism (PE)., Purpose: To compare the test characteristics of gestalt (a physician's unstructured estimate) and clinical decision rules for evaluating adults with suspected PE and assess the failure rate of gestalt and rules when used in combination with d-dimer testing., Data Sources: Articles in MEDLINE and EMBASE in English, French, German, Italian, Spanish, or Dutch that were published between 1966 and June 2011., Study Selection: 3 reviewers, working in pairs, selected prospective studies in consecutive patients suspected of having PE. Studies had to estimate the probability of PE by using gestalt or a decision rule and verify the diagnosis by using an appropriate reference standard., Data Extraction: Data on study characteristics, test performance, and prevalence were extracted. Reviewers constructed 2 × 2 tables and assessed the methodological quality of the studies., Data Synthesis: 52 studies, comprising 55 268 patients, were selected. Meta-analysis was performed on studies that used gestalt (15 studies; sensitivity, 0.85; specificity, 0.51), the Wells rule with a cutoff value less than 2 (19 studies; sensitivity, 0.84; specificity, 0.58) or 4 or less (11 studies; sensitivity, 0.60; specificity, 0.80), the Geneva rule (5 studies; sensitivity, 0.84; specificity, 0.50), and the revised Geneva rule (4 studies; sensitivity, 0.91; specificity, 0.37). An increased prevalence of PE was associated with higher sensitivity and lower specificity. Combining a decision rule or gestalt with d-dimer testing seemed safe for all strategies, except when the less-sensitive Wells rule (cutoff value ≤4) was combined with less-sensitive qualitative d-dimer testing., Limitations: Studies had substantial heterogeneity due to prevalence of PE and differences in threshold. Many studies (63%) had potential bias due to differential disease verification., Conclusion: Clinical decision rules and gestalt can safely exclude PE when combined with sensitive d-dimer testing. The authors recommend standardized rules because gestalt has lower specificity, but the choice of a particular rule and d-dimer test depend on both prevalence and setting.

Purpose: The Wells rule is widely used for clinical assessment of patients with suspected deep vein thrombosis (DVT), especially in the secondary care setting. Recently a new clinical decision rule for primary care patients (the primary care rule) has been proposed, because the Wells rule is not sufficient to rule out DVT in this setting. The objective was to compare the ability of both rules to safely rule out DVT and to efficiently reduce the number of referrals for leg ultrasound investigation that would result in a negative finding., Methods: Family physicians collected data on 1,086 patients to calculate the scores for both decision rules before leg ultrasonography was performed. In all patients D-dimer (dimerized plasmin fragment D) testing was performed using a rapid point-of-care assay. Patients were stratified into risk categories defined by each rule and the D-dimer result. Outcomes were DVT (diagnosed by ultrasonography) and venous thromboembolic complications or death caused by a possible thromboembolic event during a 90-day follow-up period. We calculated the differences between the 2 rules in the number of missed diagnoses and the proportions of patients that needed ultrasound testing., Results: Data from 1,002 eligible patients were used for this analysis. A venous thromboembolic event occurred during follow-up in 7 patients with a low score and negative D-dimer finding, both with the Wells rule (7 of 447; 1.6%; 95% confidence interval [CI], 0.7%-3.3% ) and the primary care rule (7 of 495; 1.4%; 95% CI, 0.6%-3.0%). Using the Wells rule, 447 patients (45%) would not need referral for further testing compared with 495 patients (49%) when using the primary care rule (McNemar P <.001)., Conclusions: In primary care, suspected DVT can safely be ruled out using either of the 2 rules in combination with a point-of-care D-dimer test. Both rules can reduce unnecessary referrals for compression ultrasonography by about 50%, though the primary care rule reduces it slightly more.

Background: Referral for ultrasound testing in all patients suspected of DVT is inefficient, because 80-90% have no DVT., Objective: To assess the incremental cost-effectiveness of a diagnostic strategy to select patients at first presentation in primary care based on a point of care D-dimer test combined with a clinical decision rule (AMUSE strategy), compared with hospital-based strategies., Patients/methods: A Markov-type cost-effectiveness model with a societal perspective and a 5-year time horizon was used to compare the AMUSE strategy with hospital-based strategies. Data were derived from the AMUSE study (2005-2007), the literature, and a direct survey of costs (2005-2007)., Results of Base-Case Analysis: Adherence to the AMUSE strategy on average results in savings of euro138 ($185) per patient at the expense of a very small health loss (0.002 QALYs) compared with the best hospital strategy. The iCER is euro55 753($74 848). The cost-effectiveness acceptability curves show that the AMUSE strategy has the highest probability of being cost-effective., Results of Sensitivity Analysis: Results are sensitive to decreases in sensitivity of the diagnostic strategy, but are not sensitive to increase in age (range 30-80), the costs for health states, and events., Conclusion: A diagnostic management strategy based on a clinical decision rule and a point of care D-dimer assay to exclude DVT in primary care is not only safe, but also cost-effective as compared with hospital-based strategies.

Background: Up to 90% of patients referred for ultrasonography with suspected deep venous thrombosis (DVT) of the leg do not have the disease., Objective: To evaluate the safety and efficiency of using a clinical decision rule that includes a point-of-care d-dimer assay at initial presentation in primary care to exclude DVT., Design: A prospective management study., Setting: Approximately 300 primary care practices in 3 regions of the Netherlands (Amsterdam, Maastricht, and Utrecht)., Patients: 1028 consecutive patients with clinically suspected DVT., Intervention: Patients were managed on the basis of the result of the clinical decision rule, which included a d-dimer result. Patients with a score of 3 or less were not referred for ultrasonography and received no anticoagulant treatment; patients with a score of 4 or more were referred for ultrasonography., Measurements: The primary outcome was symptomatic, objectively confirmed, venous thromboembolism during 3-month follow-up., Results: The mean age of the 1028 study patients was 58 years, and 37% of patients were men. A valid score was obtained in 1002 patients (98%). In 500 patients (49%), with a score of 3 or less, 7 developed venous thromboembolism within 3 months (incidence, 1.4% [95% CI, 0.6% to 2.9%]). A total of 502 patients (49%) had a score of 4 or more; 3 did not have ultrasonography. Ultrasonography showed DVT in 125 patients (25%), for an overall prevalence in evaluable patients of 13% (125 of 1002). Of the 374 patients who had normal ultrasonography results, 4 developed venous thromboembolism within 3 months (1.1% [CI, 0.3% to 2.7%])., Limitation: The study lacked a randomized design and relied on clinical follow-up to detect missed thrombotic disease., Conclusion: A diagnostic management strategy in primary care by using a simple clinical decision rule and a point-of-care d-dimer assay reduces the need for referral to secondary care of patients with clinically suspected DVT by almost 50% and is associated with a low risk for subsequent venous thromboembolic events., Funding: The Netherlands Organization for Scientific Research.

Objective: To compare polytomous and dichotomous logistic regression analyses in diagnosing serious bacterial infections (SBIs) in children with fever without apparent source (FWS)., Study Design and Setting: We analyzed data of 595 children aged 1-36 months, who attended the emergency department with fever without source. Outcome categories were SBI, subdivided in pneumonia and other-SBI (OSBI), and non-SBI. Potential predictors were selected based on previous studies and literature. Four models were developed: a polytomous model, estimating probabilities for three diagnostic categories simultaneously; two sequential dichotomous models, which differed in variable selection, discriminating SBI and non-SBI in step 1, and pneumonia and OSBI in step 2; and model 4, where each outcome category was opposed to the other two. The models were compared with respect to the area under the receiver-operating characteristic curve (AUC) for each of the three outcome categories and to the variable selection., Results: Small differences were found in the variables that were selected in the polytomous and dichotomous models. The AUCs of the three outcome categories were similar for each modeling strategy., Conclusion: A polytomous logistic regression analysis did not outperform sequential and single application of dichotomous logistic regression analyses in diagnosing SBIs in children with FWS.

Objective: Physicians commonly consider the presence of all differential diagnoses simultaneously. Polytomous logistic regression modeling allows for simultaneous estimation of the probability of multiple diagnoses. We discuss and (empirically) illustrate the value of this method for diagnostic research., Study Design and Setting: We used data from a study on the diagnosis of residual retroperitoneal mass histology in patients presenting with nonseminomatous testicular germ cell tumor. The differential diagnoses include benign tissue, mature teratoma, and viable cancer. Probabilities of each diagnosis were estimated with a polytomous logistic regression model and compared with the probabilities estimated from two consecutive dichotomous logistic regression models., Results: We provide interpretations of the odds ratios derived from the polytomous regression model and present a simple score chart to facilitate calculation of predicted probabilities from the polytomous model. For both modeling methods, we show the calibration plots and receiver operating characteristics curve (ROC) areas comparing each diagnostic outcome category with the other two. The ROC areas for benign tissue, mature teratoma, and viable cancer were similar for both modeling methods, 0.83 (95% confidence interval [CI]=0.80-0.85) vs. 0.83 (95% CI=0.80-0.85), 0.78 (95% CI=0.75-0.81) vs. 0.78 (95% CI=0.75-0.81), and 0.66 (95% CI=0.61-0.71) vs. 0.64 (95% CI=0.59-0.69), for polytomous and dichotomous regression models, respectively., Conclusion: Polytomous logistic regression is a useful technique to simultaneously model predicted probabilities of multiple diagnostic outcome categories. The performance of a polytomous prediction model can be assessed similarly to a dichotomous logistic regression model, and predictions by a polytomous model can be made with a user-friendly method. Because the simultaneous consideration of the presence of multiple (differential) conditions serves clinical practice better than consideration of the presence of only one target condition, polytomous logistic regression could be applied more often in diagnostic research.

Unlabelled: The aim of this study was to design a clinical rule to predict the presence of a serious bacterial infection in children with fever without apparent source. Information was collected from the records of children aged 1-36 mo who attended the paediatric emergency department because of fever without source (temperature > or = 38 degrees C and no apparent source found after evaluation by a general practitioner or history by a paediatrician). Serious bacterial infection included bacterial meningitis, sepsis, bacteraemia, pneumonia, urinary tract infection, bacterial gastroenteritis, osteomyelitis and ethmoiditis. Using multivariate logistic regression and the area under the receiver operating characteristic curve (ROC area), the diagnostic value of predictors for serious bacterial infection was judged, resulting in a risk stratification. Twenty-five percent of the 231 patients enrolled in the study (mean age 1.1 y) had a serious bacterial infection. Independent predictors from history and examination included duration of fever, poor micturition, vomiting, age, temperature < 36.7 degrees C or > or = 40 degrees C at examination, chest-wall retractions and poor peripheral circulation (ROC area: 0.75). Independent predictors from laboratory tests were white blood cell count, serum C-reactive protein and the presence of >70 white blood cells in urinalysis (ROC area: 0.83). The risk stratification for serious bacterial infection ranged from 6% to 92%., Conclusion: The probability of a serious bacterial infection in the individual patient with fever without source can be estimated more precisely by using a limited number of symptoms, signs and laboratory tests.

Diagnostic tests are often evaluated by comparison of the areas under receiver operating characteristic (ROC) curves. In this study the authors compared this approach with a more direct method that takes into account consequences of a diagnosis. Data from a prospective study of diagnosis of pulmonary embolism were used for a motivating example. Using multivariable logistic regression analysis, three diagnostic models were built and compared based on their ROC curves. Although model 1 (0.706) and model 2 (0.702) had the same ROC-curve area, they performed differently when risks and benefits of subsequent decisions were considered by applying the treatment probability threshold. Models 1 and 3 (0.611) had substantially different ROC-curve areas but performed similarly taking into account the therapeutic consequences. This demonstrates that comparison of diagnostic tests using the areas under the ROC curves may lead to erroneous conclusions about therapeutic usefulness. To correspond to daily practice, it would be more appropriate to also consider the clinical implications in evaluating diagnostic tests. This is made feasible by explicit definition and application of a treatment threshold.