Decision curve analysis revisited: overall net benefit, relationships to ROC curve analysis, and application to case-control studies.

ABSTRACT: BACKGROUND: Decision curve analysis has been introduced as a method to evaluate prediction models in terms of their clinical consequences if used for a binary classification of subjects into a group who should and into a group who should not be treated. The key concept for this type of evaluation is the "net benefit", a concept borrowed from utility theory. METHODS: We recall the foundations of decision curve analysis and discuss some new aspects. First, we stress the formal distinction between the net benefit for the treated and for the untreated and define the concept of the "overall net benefit". Next, we revisit the important distinction between the concept of accuracy, as typically assessed using the Youden index and a receiver operating characteristic (ROC) analysis, and the concept of utility of a prediction model, as assessed using decision curve analysis. Finally, we provide an explicit implementation of decision curve analysis to be applied in the context of case-control studies. RESULTS: We show that the overall net benefit, which combines the net benefit for the treated and the untreated, is a natural alternative to the benefit achieved by a model, being invariant with respect to the coding of the outcome, and conveying a more comprehensive picture of the situation. Further, within the framework of decision curve analysis, we illustrate the important difference between the accuracy and the utility of a model, demonstrating how poor an accurate model may be in terms of its net benefit. Eventually, we expose that the application of decision curve analysis to case-control studies, where an accurate estimate of the true prevalence of a disease cannot be obtained from the data, is achieved with a few modifications to the original calculation procedure. CONCLUSIONS: We present several interrelated extensions to decision curve analysis that will both facilitate its interpretation and broaden its potential area of application.

Prospective validation of the Pulmonary Embolism Severity Index. A clinical prognostic model for pulmonary embolism

Practice guidelines recommend outpatient care for selected patients with non-massive pulmonary embolism (PE), but fail to specify how these low-risk patients should be identified. Using data from U.S. patients, we previously derived the Pulmonary Embolism Severity Index (PESI), a prediction rule that risk stratifies patients with PE. We sought to validate the PESI in a European patient cohort. We prospectively validated the PESI in patients with PE diagnosed at six emergency departments in three European countries. We used baseline data for the rule's 11 prognostic variables to stratify patients into five risk classes (I-V) of increasing probability of mortality. The outcome was overall mortality at 90 days after presentation. To assess the accuracy of the PESI to predict mortality, we estimated the sensitivity, specificity, and predictive values for low- (risk classes I/II) versus higher-risk patients (risk classes III-V), and the discriminatory power using the area under the receiver operating characteristic (ROC) curve. Among 357 patients with PE, overall mortality was 5.9%, ranging from 0% in class I to 17.9% in class V. The 186 (52%) low-risk patients had an overall mortality of 1.1% (95% confidence interval [CI]: 0.1-3.8%) compared to 11.1% (95% CI: 6.8-16.8%) in the 171 (48%) higher-risk patients. The PESI had a high sensitivity (91%, 95% CI: 71-97%) and a negative predictive value (99%, 95% CI: 96-100%) for predicting mortality. The area under the ROC curve was 0.78 (95% CI: 0.70-0.86). The PESI reliably identifies patients with PE who are at low risk of death and who are potential candidates for outpatient care. The PESI may help physicians make more rational decisions about hospitalization for patients with PE.

Prospective, multicenter validation of prediction scores for major bleeding in elderly patients with venous thromboembolism.

BACKGROUND: The Outpatient Bleeding Risk Index (OBRI) and the Kuijer, RIETE and Kearon scores are clinical prognostic scores for bleeding in patients receiving oral anticoagulants for venous thromboembolism (VTE). We prospectively compared the performance of these scores in elderly patients with VTE. METHODS: In a prospective multicenter Swiss cohort study, we studied 663 patients aged ≥ 65 years with acute VTE. The outcome was a first major bleeding at 90 days. We classified patients into three categories of bleeding risk (low, intermediate and high) according to each score and dichotomized patients as high vs. low or intermediate risk. We calculated the area under the receiver-operating characteristic (ROC) curve, positive predictive values and likelihood ratios for each score. RESULTS: Overall, 28 out of 663 patients (4.2%, 95% confidence interval [CI] 2.8-6.0%) had a first major bleeding within 90 days. According to different scores, the rate of major bleeding varied from 1.9% to 2.1% in low-risk, from 4.2% to 5.0% in intermediate-risk and from 3.1% to 6.6% in high-risk patients. The discriminative power of the scores was poor to moderate, with areas under the ROC curve ranging from 0.49 to 0.60 (P = 0.21). The positive predictive values and positive likelihood ratios were low and varied from 3.1% to 6.6% and from 0.72 to 1.59, respectively. CONCLUSION: In elderly patients with VTE, existing bleeding risk scores do not have sufficient accuracy and power to discriminate between patients with VTE who are at a high risk of short-term major bleeding and those who are not.

Pancreatic stone protein as a novel marker for neonatal sepsis.

PURPOSE: Early-onset sepsis (EOS) is one of the main causes for the admission of newborns to the neonatal intensive care unit. However, traditional infection markers are poor diagnostic markers of EOS. Pancreatic stone protein (PSP) is a promising sepsis marker in adults. The aim of this study was to investigate whether determining PSP improves the diagnosis of EOS in comparison with other infection markers. METHODS: This was a prospective multicentre study involving 137 infants with a gestational age of >34 weeks who were admitted with suspected EOS. PSP, procalcitonin (PCT), soluble human triggering receptor expressed on myeloid cells-1 (sTREM-1), macrophage migration inhibitory factor (MIF) and C-reactive protein (CRP) were measured at admission. Receiver-operating characteristic (ROC) curve analysis was performed. RESULTS: The level of PSP in infected infants was significantly higher than that in uninfected ones (median 11.3 vs. 7.5 ng/ml, respectively; p = 0.001). The ROC area under the curve was 0.69 [95 % confidence interval (CI) 0.59-0.80; p < 0.001] for PSP, 0.77 (95 % CI 0.66-0.87; p < 0.001) for PCT, 0.66 (95 % CI 0.55-0.77; p = 0.006) for CRP, 0.62 (0.51-0.73; p = 0.055) for sTREM-1 and 0.54 (0.41-0.67; p = 0.54) for MIF. PSP independently of PCT predicted EOS (p < 0.001), and the use of both markers concomitantly significantly increased the ability to diagnose EOS. A bioscore combining PSP (>9 ng/ml) and PCT (>2 ng/ml) was the best predictor of EOS (0.83; 95 % CI 0.74-0.93; p < 0.001) and resulted in a negative predictive value of 100 % and a positive predictive value of 71 %. CONCLUSIONS: In this prospective study, the diagnostic performance of PSP and PCT was superior to that of traditional markers and a combination bioscore improved the diagnosis of sepsis. Our findings suggest that PSP is a valuable biomarker in combination with PCT in EOS.

Performance of blood pressure-to-height ratio at a single screening visit for the identification of hypertension in children.

BACKGROUND: The diagnosis of hypertension in children is difficult because of the multiple sex-, age-, and height-specific thresholds to define elevated blood pressure (BP). Blood pressure-to-height ratio (BPHR) has been proposed to facilitate the identification of elevated BP in children. OBJECTIVE: We assessed the performance of BPHR at a single screening visit to identify children with hypertension that is sustained elevated BP. METHOD: In a school-based study conducted in Switzerland, BP was measured at up to three visits in 5207 children. Children had hypertension if BP was elevated at the three visits. Sensitivity, specificity, negative predictive value (NPV), and positive predictive value (PPV) for the identification of hypertension were assessed for different thresholds of BPHR. The ability of BPHR at a single screening visit to discriminate children with and without hypertension was evaluated with receiver operating characteristic (ROC) curve analyses. RESULTS: The prevalence of systolic/diastolic hypertension was 2.2%. Systolic BPHR had a better performance to identify hypertension compared with diastolic BPHR (area under the ROC curve: 0.95 vs. 0.84). The highest performance was obtained with a systolic BPHR threshold set at 0.80 mmHg/cm (sensitivity: 98%; specificity: 85%; PPV: 12%; and NPV: 100%) and a diastolic BPHR threshold set at 0.45 mmHg/cm (sensitivity: 79%; specificity: 70%; PPV: 5%; and NPV: 99%). The PPV was higher among tall or overweight children. CONCLUSION: BPHR at a single screening visit had a high performance to identify hypertension in children, although the low prevalence of hypertension led to a low PPV.

Event-related potential analyses via single-subject topographic classification.

Introduction: Responses to external stimuli are typically investigated by averaging peri-stimulus electroencephalography (EEG) epochs in order to derive event-related potentials (ERPs) across the electrode montage, under the assumption that signals that are related to the external stimulus are fixed in time across trials. We demonstrate the applicability of a single-trial model based on patterns of scalp topographies (De Lucia et al, 2007) that can be used for ERP analysis at the single-subject level. The model is able to classify new trials (or groups of trials) with minimal a priori hypotheses, using information derived from a training dataset. The features used for the classification (the topography of responses and their latency) can be neurophysiologically interpreted, because a difference in scalp topography indicates a different configuration of brain generators. An above chance classification accuracy on test datasets implicitly demonstrates the suitability of this model for EEG data. Methods: The data analyzed in this study were acquired from two separate visual evoked potential (VEP) experiments. The first entailed passive presentation of checkerboard stimuli to each of the four visual quadrants (hereafter, "Checkerboard Experiment") (Plomp et al, submitted). The second entailed active discrimination of novel versus repeated line drawings of common objects (hereafter, "Priming Experiment") (Murray et al, 2004). Four subjects per experiment were analyzed, using approx. 200 trials per experimental condition. These trials were randomly separated in training (90%) and testing (10%) datasets in 10 independent shuffles. In order to perform the ERP analysis we estimated the statistical distribution of voltage topographies by a Mixture of Gaussians (MofGs), which reduces our original dataset to a small number of representative voltage topographies. We then evaluated statistically the degree of presence of these template maps across trials and whether and when this was different across experimental conditions. Based on these differences, single-trials or sets of a few single-trials were classified as belonging to one or the other experimental condition. Classification performance was assessed using the Receiver Operating Characteristic (ROC) curve. Results: For the Checkerboard Experiment contrasts entailed left vs. right visual field presentations for upper and lower quadrants, separately. The average posterior probabilities, indicating the presence of the computed template maps in time and across trials revealed significant differences starting at ~60-70 ms post-stimulus. The average ROC curve area across all four subjects was 0.80 and 0.85 for upper and lower quadrants, respectively and was in all cases significantly higher than chance (unpaired t-test, p<0.0001). In the Priming Experiment, we contrasted initial versus repeated presentations of visual object stimuli. Their posterior probabilities revealed significant differences, which started at 250ms post-stimulus onset. The classification accuracy rates with single-trial test data were at chance level. We therefore considered sub-averages based on five single trials. We found that for three out of four subjects' classification rates were significantly above chance level (unpaired t-test, p<0.0001). Conclusions: The main advantage of the present approach is that it is based on topographic features that are readily interpretable along neurophysiologic lines. As these maps were previously normalized by the overall strength of the field potential on the scalp, a change in their presence across trials and between conditions forcibly reflects a change in the underlying generator configurations. The temporal periods of statistical difference between conditions were estimated for each training dataset for ten shuffles of the data. Across the ten shuffles and in both experiments, we observed a high level of consistency in the temporal periods over which the two conditions differed. With this method we are able to analyze ERPs at the single-subject level providing a novel tool to compare normal electrophysiological responses versus single cases that cannot be considered part of any cohort of subjects. This aspect promises to have a strong impact on both basic and clinical research.

Development and validation of a clinical prediction rule for chest wall syndrome in primary care.

ABSTRACT: BACKGROUND: Chest wall syndrome (CWS), the main cause of chest pain in primary care practice, is most often an exclusion diagnosis. We developed and evaluated a clinical prediction rule for CWS. METHODS: Data from a multicenter clinical cohort of consecutive primary care patients with chest pain were used (59 general practitioners, 672 patients). A final diagnosis was determined after 12 months of follow-up. We used the literature and bivariate analyses to identify candidate predictors, and multivariate logistic regression was used to develop a clinical prediction rule for CWS. We used data from a German cohort (n = 1212) for external validation. RESULTS: From bivariate analyses, we identified six variables characterizing CWS: thoracic pain (neither retrosternal nor oppressive), stabbing, well localized pain, no history of coronary heart disease, absence of general practitioner's concern, and pain reproducible by palpation. This last variable accounted for 2 points in the clinical prediction rule, the others for 1 point each; the total score ranged from 0 to 7 points. The area under the receiver operating characteristic (ROC) curve was 0.80 (95% confidence interval 0.76-0.83) in the derivation cohort (specificity: 89%; sensitivity: 45%; cut-off set at 6 points). Among all patients presenting CWS (n = 284), 71% (n = 201) had a pain reproducible by palpation and 45% (n = 127) were correctly diagnosed. For a subset (n = 43) of these correctly classified CWS patients, 65 additional investigations (30 electrocardiograms, 16 thoracic radiographies, 10 laboratory tests, eight specialist referrals, one thoracic computed tomography) had been performed to achieve diagnosis. False positives (n = 41) included three patients with stable angina (1.8% of all positives). External validation revealed the ROC curve to be 0.76 (95% confidence interval 0.73-0.79) with a sensitivity of 22% and a specificity of 93%. CONCLUSIONS: This CWS score offers a useful complement to the usual CWS exclusion diagnosing process. Indeed, for the 127 patients presenting CWS and correctly classified by our clinical prediction rule, 65 additional tests and exams could have been avoided. However, the reproduction of chest pain by palpation, the most important characteristic to diagnose CWS, is not pathognomonic.

Urinary low-molecular-weight protein excretion in pediatric idiopathic nephrotic syndrome.

BACKGROUND: Minimal change disease (MCD) and focal segmental glomerulosclerosis (FSGS) are the most common causes of idiopathic nephrotic syndrome (INS). We have evaluated the reliability of urinary neutrophil-gelatinase-associated lipocalin (uNGAL), urinary alpha1-microglobulin (uα1M) and urinary N-acetyl-beta-D-glucosaminidase (uβNAG) as markers for differentiating MCD from FSGS. We have also evaluated whether these proteins are associated to INS relapses or to glomerular filtration rate (GFR). METHODS: The patient cohort comprised 35 children with MCD and nine with FSGS; 19 healthy age-matched children were included in the study as controls. Of the 35 patients, 28 were in remission (21 MCD, 7 FSGS) and 16 were in relapse (14 MCD, 2 FSGS). The prognostic accuracies of these proteins were assessed by receiver operating characteristic (ROC) curve analyses. RESULTS: The level of uNGAL, indexed or not to urinary creatinine (uCreat), was significantly different between children with INS and healthy children (p = 0.02), between healthy children and those with FSGS (p = 0.007) and between children with MCD and those with FSGS (p = 0.01). It was not significantly correlated to proteinuria or GFR levels. The ROC curve analysis showed that a cut-off value of 17 ng/mg for the uNGAL/uCreat ratio could be used to distinguish MCD from FSGS with a sensitivity of 0.77 and specificity of 0.78. uβNAG was not significantly different in patients with MCD and those with FSGS (p = 0.86). Only uα1M, indexed or not to uCreat, was significantly (p < 0.001) higher for patients in relapse compared to those in remission. CONCLUSIONS: Our results indicate that in our patient cohort uNGAL was a reliable biomarker for differentiating MCD from FSGS independently of proteinuria or GFR levels.

Ability of different adiposity indicators to identify children with elevated blood pressure.

OBJECTIVE: Body composition measured by dual-energy X-ray absorptiometry (DXA) is believed to be superior to crude measures such as BMI or waist circumference (WC) to assess health risks associated with adiposity in adults. We compared the ability of BMI, WC, waist-to-height ratio (WHtR), percentage body fat from skinfold thickness, and measures of total and central fat assessed by DXA to identify children with elevated blood pressure (BP). STUDY DESIGN: The QUALITY Study follows 630 Caucasian families (father, mother, and child originally aged 8-10 years). BP, height, weight, WC, and skinfold thickness were measured according to standardized protocols. Elevated BP was defined as systolic or diastolic BP at least 90th age, sex, and height-specific percentile. Total and central fat were determined with DXA. The area under the receiver operating characteristic (ROC) curve (AUC) statistic was computed from logistic models that adjusted for age, sex, height, Tanner stage, and physical activity. RESULTS: All adiposity indicators were highly correlated. WC and WHtR did not show superior ability over BMI to identify children with elevated SBP (P = 0.421 and 0.473). Measures of total and central fat from DXA did not show an improved ability over BMI or WC to identify children with elevated SBP (P = 0.325-0.662). CONCLUSION: Results support the use of BMI in clinical and public health settings, at least in this age group. As all indicators had a limited ability to identify children with elevated BP, results also support measurement of BP in all children of this age independent of a weight status.

The Atherosclerosis Burden Score (ABS): a Convenient Ultrasound-Based Score of Peripheral Atherosclerosis for Coronary Artery Disease Prediction.

Ultrasonographic detection of subclinical atherosclerosis improves cardiovascular risk stratification, but uncertainty persists about the most discriminative method to apply. In this study, we found that the "atherosclerosis burden score (ABS)", a novel straightforward ultrasonographic score that sums the number of carotid and femoral arterial bifurcations with plaques, significantly outperformed common carotid intima-media thickness, carotid mean/maximal thickness, and carotid/femoral plaque scores for the detection of coronary artery disease (CAD) (receiver operating characteristic (ROC) curve area under the curve (AUC) = 0.79; P = 0.027 to <0.001 with the other five US endpoints) in 203 patients undergoing coronary angiography. ABS was also more correlated with CAD extension (R = 0.55; P < 0.001). Furthermore, in a second group of 1128 patients without cardiovascular disease, ABS was weakly correlated with the European Society of Cardiology chart risk categories (R (2) = 0.21), indicating that ABS provided information beyond usual cardiovascular risk factor-based risk stratification. Pending prospective studies on hard cardiovascular endpoints, ABS appears as a promising tool in primary prevention.

A genetic risk score combining 32 SNPs is associated with body mass index and improves obesity prediction in people with major depressive disorder.

BACKGROUND: Obesity is strongly associated with major depressive disorder (MDD) and various other diseases. Genome-wide association studies have identified multiple risk loci robustly associated with body mass index (BMI). In this study, we aimed to investigate whether a genetic risk score (GRS) combining multiple BMI risk loci might have utility in prediction of obesity in patients with MDD. METHODS: Linear and logistic regression models were conducted to predict BMI and obesity, respectively, in three independent large case-control studies of major depression (Radiant, GSK-Munich, PsyCoLaus). The analyses were first performed in the whole sample and then separately in depressed cases and controls. An unweighted GRS was calculated by summation of the number of risk alleles. A weighted GRS was calculated as the sum of risk alleles at each locus multiplied by their effect sizes. Receiver operating characteristic (ROC) analysis was used to compare the discriminatory ability of predictors of obesity. RESULTS: In the discovery phase, a total of 2,521 participants (1,895 depressed patients and 626 controls) were included from the Radiant study. Both unweighted and weighted GRS were highly associated with BMI (P <0.001) but explained only a modest amount of variance. Adding 'traditional' risk factors to GRS significantly improved the predictive ability with the area under the curve (AUC) in the ROC analysis, increasing from 0.58 to 0.66 (95% CI, 0.62-0.68; χ(2) = 27.68; P <0.0001). Although there was no formal evidence of interaction between depression status and GRS, there was further improvement in AUC in the ROC analysis when depression status was added to the model (AUC = 0.71; 95% CI, 0.68-0.73; χ(2) = 28.64; P <0.0001). We further found that the GRS accounted for more variance of BMI in depressed patients than in healthy controls. Again, GRS discriminated obesity better in depressed patients compared to healthy controls. We later replicated these analyses in two independent samples (GSK-Munich and PsyCoLaus) and found similar results. CONCLUSIONS: A GRS proved to be a highly significant predictor of obesity in people with MDD but accounted for only modest amount of variance. Nevertheless, as more risk loci are identified, combining a GRS approach with information on non-genetic risk factors could become a useful strategy in identifying MDD patients at higher risk of developing obesity.

Test of four colon cancer risk-scores in formalin fixed paraffin embedded microarray gene expression data.

BACKGROUND: Prognosis prediction for resected primary colon cancer is based on the T-stage Node Metastasis (TNM) staging system. We investigated if four well-documented gene expression risk scores can improve patient stratification. METHODS: Microarray-based versions of risk-scores were applied to a large independent cohort of 688 stage II/III tumors from the PETACC-3 trial. Prognostic value for relapse-free survival (RFS), survival after relapse (SAR), and overall survival (OS) was assessed by regression analysis. To assess improvement over a reference, prognostic model was assessed with the area under curve (AUC) of receiver operating characteristic (ROC) curves. All statistical tests were two-sided, except the AUC increase. RESULTS: All four risk scores (RSs) showed a statistically significant association (single-test, P < .0167) with OS or RFS in univariate models, but with HRs below 1.38 per interquartile range. Three scores were predictors of shorter RFS, one of shorter SAR. Each RS could only marginally improve an RFS or OS model with the known factors T-stage, N-stage, and microsatellite instability (MSI) status (AUC gains < 0.025 units). The pairwise interscore discordance was never high (maximal Spearman correlation = 0.563) A combined score showed a trend to higher prognostic value and higher AUC increase for OS (HR = 1.74, 95% confidence interval [CI] = 1.44 to 2.10, P < .001, AUC from 0.6918 to 0.7321) and RFS (HR = 1.56, 95% CI = 1.33 to 1.84, P < .001, AUC from 0.6723 to 0.6945) than any single score. CONCLUSIONS: The four tested gene expression-based risk scores provide prognostic information but contribute only marginally to improving models based on established risk factors. A combination of the risk scores might provide more robust information. Predictors of RFS and SAR might need to be different.

A multi-temporal approach to model endangered species distribution in Europe: the case of the Eurasian otter in Italy

We analysed the relationship between changes in land cover patterns and the Eurasian otter occurrence over the course of about 20 years (1985-2006) using multi-temporal Species Distribution Models (SDMs). The study area includes five river catchments covering most of the otter's Italian range. Land cover and topographic data were used as proxies of the ecological requirements of the otter within a 300-m buffer around river courses. We used species presence, pseudo-absence data, and environmental predictors to build past (1985) and current (2006) SDMs by applying an ensemble procedure through the BIOMOD modelling package. The performance of each model was evaluated by measuring the area under the curve (AUC) of the receiver-operating characteristic (ROC). Multi-temporal analyses of species distribution and land cover maps were performed by comparing the maps produced for 1985 and 2006. The ensemble procedure provided a good overall modelling accuracy, revealing that elevation and slope affected the otter's distribution in the past; in contrast, land cover predictors, such as cultivations and forests, were more important in the present period. During the transition period, 20.5% of the area became suitable, with 76% of the new otter presence data being located in these newly available areas. The multi-temporal analysis suggested that the quality of otter habitat improved in the last 20 years owing to the expansion of forests and to the reduction of cultivated fields in riparian belts. The evidence presented here stresses the great potential of riverine habitat restoration and environmental management for the future expansion of the otter in Italy

CSF lactate for accurate diagnosis of community-acquired bacterial meningitis.

CSF lactate measurement is recommended when nosocomial meningitis is suspected, but its value in community-acquired bacterial meningitis is controversial. We evaluated the diagnostic performance of lactate and other CSF parameters in a prospective cohort of adult patients with acute meningitis. Diagnostic accuracy of lactate and other CSF parameters in patients with microbiologically documented episodes was assessed by receiver operating characteristic (ROC) curves. The cut-offs with the best diagnostic performance were determined. Forty-five of 61 patients (74%) had a documented bacterial (n = 18; S. pneumoniae, 11; N. meningitidis, 5; other, 2) or viral (n = 27 enterovirus, 21; VZV, 3; other, 3) etiology. CSF parameters were significantly different in bacterial vs. viral meningitis, respectively (p < 0.001 for all comparisons): white cell count (median 1333 vs. 143/mm(3)), proteins (median 4115 vs. 829 mg/l), CSF/blood glucose ratio (median 0.1 vs. 0.52), lactate (median 13 vs. 2.3 mmol/l). ROC curve analysis showed that CSF lactate had the highest accuracy for discriminating bacterial from viral meningitis, with a cutoff set at 3.5 mmol/l providing 100% sensitivity, specificity, PPV, NPV, and efficiency. CSF lactate had the best accuracy for discriminating bacterial from viral meningitis and should be included in the initial diagnostic workup of this condition.

Immuno-reactive proteins from Mycobacterium immunogenum useful for serodiagnosis of metalworking fluid hypersensitivity pneumonitis.

Metalworking fluid-associated hypersensitivity pneumonitis (MWF-HP) is a pulmonary disease caused by inhaling microorganisms present in the metalworking fluids used in the industrial sector. Mycobacterium immunogenum is the main etiological agent. Among the clinical, radiological and biological tools used for diagnosis, serological tests are important. The aim of this study was to identify immunogenic proteins in M. immunogenum and to use recombinant antigens for serological diagnosis of MWF-HP. Immunogenic proteins were detected by two-dimensional Western blot and candidate proteins were identified by mass spectrometry. Recombinant antigens were expressed in Escherichia coli and tested by enzyme-linked immunosorbent assay (ELISA) with the sera of 14 subjects with MWF-HP and 12 asymptomatic controls exposed to M. immunogenum. From the 350 spots visualized by two-dimensional gel electrophoresis with M. immunogenum extract, 6 immunogenic proteins were selected to be expressed as recombinant antigens. Acyl-CoA dehydrogenase antigen allowed for the best discrimination of MWF-HP cases against controls with an area under the receiver operating characteristics (ROC) curve of 0.930 (95% CI=0.820-1), a sensitivity of 100% and a specificity of 83% for the optimum threshold. Other recombinant antigens correspond to acyl-CoA dehydrogenase FadE, cytosol aminopeptidase, dihydrolipoyl dehydrogenase, serine hydroxymethyltransferase and superoxide dismutase. This is the first time that recombinant antigens have been used for the serodiagnosis of hypersensitivity pneumonitis. The availability of recombinant antigens makes it possible to develop standardized serological tests which in turn could simplify diagnosis, thus making it less invasive.