Diagnostic and prognostic research最新文献

A lack of biomarkers that detect drug-induced liver injury (DILI) accurately continues to hinder early- and late-stage drug development and remains a challenge in clinical practice. The Innovative Medicines Initiative's TransBioLine consortium comprising academic and industry partners is developing a prospective repository of deeply phenotyped cases and controls with biological samples during liver injury progression to facilitate biomarker discovery, evaluation, validation and qualification.In a nested case-control design, patients who meet one of these criteria, alanine transaminase (ALT) ≥ 5 × the upper limit of normal (ULN), alkaline phosphatase ≥ 2 × ULN or ALT ≥ 3 ULN with total bilirubin > 2 × ULN, are enrolled. After completed clinical investigations, Roussel Uclaf Causality Assessment and expert panel review are used to adjudicate episodes as DILI or alternative liver diseases (acute non-DILI controls). Two blood samples are taken: at recruitment and follow-up. Sample size is as follows: 300 cases of DILI and 130 acute non-DILI controls. Additional cross-sectional cohorts (1 visit) are as follows: Healthy volunteers (n = 120), controls with chronic alcohol-related or non-alcoholic fatty liver disease (n = 100 each) and patients with psoriasis or rheumatoid arthritis (n = 100, 50 treated with methotrexate) are enrolled. Candidate biomarkers prioritised for evaluation include osteopontin, glutamate dehydrogenase, cytokeratin-18 (full length and caspase cleaved), macrophage-colony-stimulating factor 1 receptor and high mobility group protein B1 as well as bile acids, sphingolipids and microRNAs. The TransBioLine project is enabling biomarker discovery and validation that could improve detection, diagnostic accuracy and prognostication of DILI in premarketing clinical trials and for clinical healthcare application.

Background: A previous individual participant data meta-analysis (IPD-MA) of antibiotics for adults with clinically diagnosed acute rhinosinusitis (ARS) showed a marginal overall effect of antibiotics, but was unable to identify patients that are most likely to benefit from antibiotics when applying conventional (i.e. univariable or one-variable-at-a-time) subgroup analysis. We updated the systematic review and investigated whether multivariable prediction of patient-level prognosis and antibiotic treatment effect may lead to more tailored treatment assignment in adults presenting to primary care with ARS.

Methods: An IPD-MA of nine double-blind placebo-controlled trials of antibiotic treatment (n=2539) was conducted, with the probability of being cured at 8-15 days as the primary outcome. A logistic mixed effects model was developed to predict the probability of being cured based on demographic characteristics, signs and symptoms, and antibiotic treatment assignment. Predictive performance was quantified based on internal-external cross-validation in terms of calibration and discrimination performance, overall model fit, and the accuracy of individual predictions.

Results: Results indicate that the prognosis with respect to risk of cure could not be reliably predicted (c-statistic 0.58 and Brier score 0.24). Similarly, patient-level treatment effect predictions did not reliably distinguish between those that did and did not benefit from antibiotics (c-for-benefit 0.50).

Conclusions: In conclusion, multivariable prediction based on patient demographics and common signs and symptoms did not reliably predict the patient-level probability of cure and antibiotic effect in this IPD-MA. Therefore, these characteristics cannot be expected to reliably distinguish those that do and do not benefit from antibiotics in adults presenting to primary care with ARS.

Background: In a pandemic setting, it is critical to evaluate and deploy accurate diagnostic tests rapidly. This relies heavily on the sample size chosen to assess the test accuracy (e.g. sensitivity and specificity) during the diagnostic accuracy study. Too small a sample size will lead to imprecise estimates of the accuracy measures, whereas too large a sample size may delay the development process unnecessarily. This study considers use of a Bayesian method to guide sample size determination for diagnostic accuracy studies, with application to COVID-19 rapid viral detection tests. Specifically, we investigate whether utilising existing information (e.g. from preceding laboratory studies) within a Bayesian framework can reduce the required sample size, whilst maintaining test accuracy to the desired precision.

Methods: The method presented is based on the Bayesian concept of assurance which, in this context, represents the unconditional probability that a diagnostic accuracy study yields sensitivity and/or specificity intervals with the desired precision. We conduct a simulation study to evaluate the performance of this approach in a variety of COVID-19 settings, and compare it to commonly used power-based methods. An accompanying interactive web application is available, which can be used by researchers to perform the sample size calculations.

Results: Results show that the Bayesian assurance method can reduce the required sample size for COVID-19 diagnostic accuracy studies, compared to standard methods, by making better use of laboratory data, without loss of performance. Increasing the size of the laboratory study can further reduce the required sample size in the diagnostic accuracy study.

Conclusions: The method considered in this paper is an important advancement for increasing the efficiency of the evidence development pathway. It has highlighted that the trade-off between lab study sample size and diagnostic accuracy study sample size should be carefully considered, since establishing an adequate lab sample size can bring longer-term gains. Although emphasis is on its use in the COVID-19 pandemic setting, where we envisage it will have the most impact, it can be usefully applied in other clinical areas.

Background: Numerous biomarkers have been proposed for diagnosis, therapeutic, and prognosis in sepsis. Previous evaluations of the value of biomarkers for predicting mortality due to this life-threatening condition fail to address the complexity of this condition and the risk of bias associated with prognostic studies. We evaluate the predictive performance of four of these biomarkers in the prognosis of mortality through a methodologically sound evaluation.

Methods: We conducted a systematic review a systematic review and meta-analysis to determine, in critically ill adults with sepsis, whether procalcitonin (PCT), C-reactive protein (CRP), interleukin-6 (IL-6), and presepsin (sCD14) are independent prognostic factors for mortality. We searched MEDLINE, EMBASE, and the Cochrane Central Register of Controlled Trials up to March 2023. Only Phase-2 confirmatory prognostic factor studies among critically ill septic adults were included. Random effects meta-analyses pooled the prognostic association estimates.

Results: We included 60 studies (15,681 patients) with 99 biomarker assessments. Quality of the statistical analysis and reporting domains using the QUIPS tool showed high risk of bias in > 60% assessments. The biomarker measurement as a continuous variable in models adjusted by key covariates (age and severity score) for predicting mortality at 28-30 days showed a null or near to null association for basal PCT (pooled OR = 0.99, 95% CI = 0.99-1.003), CRP (OR = 1.01, 95% CI = 0.87 to 1.17), and IL-6 (OR = 1.02, 95% CI = 1.01-1.03) and sCD14 (pooled HR = 1.003, 95% CI = 1.000 to 1.006). Additional meta-analyses accounting for other prognostic covariates had similarly null findings.

Conclusion: Baseline, isolated measurement of PCT, CRP, IL-6, and sCD14 has not been shown to help predict mortality in critically ill patients with sepsis. The role of these biomarkers should be evaluated in new studies where the patient selection would be standardized and the measurement of biomarker results.

Trial registration: PROSPERO (CRD42019128790).

Background: Rapid antigen tests detecting SARS-CoV-2 were shown to be a useful tool in managing the COVID-19 pandemic. Here, we report on the results of a prospective diagnostic accuracy study of four SARS-CoV-2 rapid antigen tests in a South African setting.

Methods: Rapid antigen test evaluations were performed through drive-through testing centres in Durban, South Africa, from July to December 2021. Two evaluation studies were performed: nasal Panbio COVID-19 Ag Rapid Test Device (Abbott) was evaluated in parallel with the nasopharyngeal Espline SARS-CoV-2 Ag test (Fujirebio), followed by the evaluation of nasal RightSign COVID-19 Antigen Rapid test Cassette (Hangzhou Biotest Biotech) in parallel with the nasopharyngeal STANDARD Q COVID-19 Ag test (SD Biosensor). The Abbott RealTime SARS-CoV-2 assay was used as a reference test.

Results: Evaluation of Panbio and Espline Ag tests was performed on 494 samples (31% positivity), while the evaluation of Standard Q and RightTest Ag tests was performed on 539 samples (13.17% positivity). The overall sensitivity for all four tests ranged between 60 and 72% with excellent specificity values (> 98%). Sensitivity increased to > 80% in all tests in samples with cycle number value < 20. All four tests performed best in samples from patients presenting within the first week of symptom onset.

Conclusions: All four evaluated tests detected a majority of the cases within the first week of symptom onset with high viral load.

Background: Group A streptococcus is found in 20-40% of cases of childhood pharyngitis; the remaining cases are viral. Streptococcal pharyngitis ("strep throat") is usually treated with antibiotics, while these are not indicated in viral cases. Most guidelines recommend relying on a diagnostic test confirming the presence of group A streptococcus before prescribing antibiotics. Conventional first-line tests are rapid antigen detection tests based on throat swabs. Recently, rapid nucleic acid tests were developed; they allow the detection of elements of the genome of group A streptococcus. We hypothesize that these rapid nucleic acid tests are sensitive enough to be performed on saliva samples instead of throat swabs, which could be more convenient in practice.

Methods: This is a multicenter, prospective diagnostic accuracy study evaluating the performance of a rapid nucleic acid test for group A streptococcus (Abbott ID NOW STREP A2) in saliva, compared with a conventional pharyngeal rapid antigen detection test (EXACTO PRO STREPTATEST, lateral flow assay, comparator test), with a composite reference standard of throat culture and group A streptococcus PCR in children with pharyngitis in primary care (i.e., 27 primary care pediatricians or general practitioners). To ensure group A streptococcus is not missed, the salivary rapid nucleic acid test requires a minimally acceptable value of sensitivity (primary outcome) set at 80%. Assuming 35% of participants will have group A streptococcus, we will recruit 800 consecutive children with pharyngitis. Secondary outcomes will include difference in sensitivity between the pharyngeal rapid antigen detection test and the salivary rapid nucleic acid test; variability in sensitivity and specificity of the salivary rapid nucleic acid test with the level of McIsaac score; time to obtain the result of the salivary rapid nucleic acid test; patient, physician, and parents satisfaction; and barriers and facilitators to using rapid tests for group A streptococcus in primary care.

Ethics and dissemination: Approved by the Institutional Review Board "Comité de protection des personnes Ile de France I" (no. 2022-A00085-38). Results will be presented at international meetings and disseminated in peer-reviewed journals.

Trial registration number: ClinicalTrials.gov: NCT05521568.

Background: Clinical scores help physicians to make clinical decisions, and some are recommended by health authorities for primary care use. As an increasing number of scores are becoming available, there is a need to understand general practitioner expectations for their use in primary care. The aim of this study was to explore general practitioner opinions about using scores in general practice.

Method: This qualitative study, with a grounded theory approach, used focus groups with general practitioners recruited from their own surgeries to obtain verbatim. Two investigators performed verbatim analysis to ensure data triangulation. The verbatim was double-blind labeled for inductive categorization to conceptualize score use in general practice.

Results: Five focus groups were planned, 21 general practitioners from central France participated. Participants appreciated scores for their clinical efficacy but felt that they were difficult to use in primary care. Their opinions revolved around validity, acceptability, and feasibility. Participants have little regard for score validity, they felt many scores are difficult to accept and do not capture contextual and human elements. Participants also felt that scores are unfeasible for primary care use. There are too many, they are hard to find, and either too short or too long. They also felt that scores were complex to administer and took up time for both patient and physician. Many participants felt learned societies should choose appropriate scores.

Discussion: This study conceptualizes general practitioner opinions about score use in primary care. The participants weighed up score effectiveness with efficiency. For some participants, scores helped make decisions faster, others expressed being disappointed with the lack of patient-centeredness and limited bio-psycho-social approach.

Background: A number of recent papers have proposed methods to calculate confidence intervals and p values for net benefit used in decision curve analysis. These papers are sparse on the rationale for doing so. We aim to assess the relation between sampling variability, inference, and decision-analytic concepts.

Methods and results: We review the underlying theory of decision analysis. When we are forced into a decision, we should choose the option with the highest expected utility, irrespective of p values or uncertainty. This is in some distinction to traditional hypothesis testing, where a decision such as whether to reject a given hypothesis can be postponed. Application of inference for net benefit would generally be harmful. In particular, insisting that differences in net benefit be statistically significant would dramatically change the criteria by which we consider a prediction model to be of value. We argue instead that uncertainty related to sampling variation for net benefit should be thought of in terms of the value of further research. Decision analysis tells us which decision to make for now, but we may also want to know how much confidence we should have in that decision. If we are insufficiently confident that we are right, further research is warranted.

Conclusion: Null hypothesis testing or simple consideration of confidence intervals are of questionable value for decision curve analysis, and methods such as value of information analysis or approaches to assess the probability of benefit should be considered instead.

Prediction algorithms that quantify the expected benefit of a given treatment conditional on patient characteristics can critically inform medical decisions. Quantifying the performance of treatment benefit prediction algorithms is an active area of research. A recently proposed metric, the concordance statistic for benefit (cfb), evaluates the discriminative ability of a treatment benefit predictor by directly extending the concept of the concordance statistic from a risk model with a binary outcome to a model for treatment benefit. In this work, we scrutinize cfb on multiple fronts. Through numerical examples and theoretical developments, we show that cfb is not a proper scoring rule. We also show that it is sensitive to the unestimable correlation between counterfactual outcomes and to the definition of matched pairs. We argue that measures of statistical dispersion applied to predicted benefits do not suffer from these issues and can be an alternative metric for the discriminatory performance of treatment benefit predictors.

Background: The performance of models for binary outcomes can be described by measures such as the concordance statistic (c-statistic, area under the curve), the discrimination slope, or the Brier score. At internal validation, data resampling techniques, e.g., cross-validation, are frequently employed to correct for optimism in these model performance criteria. Especially with small samples or rare events, leave-one-out cross-validation is a popular choice.

Methods: Using simulations and a real data example, we compared the effect of different resampling techniques on the estimation of c-statistics, discrimination slopes, and Brier scores for three estimators of logistic regression models, including the maximum likelihood and two maximum penalized likelihood estimators.

Results: Our simulation study confirms earlier studies reporting that leave-one-out cross-validated c-statistics can be strongly biased towards zero. In addition, our study reveals that this bias is even more pronounced for model estimators shrinking estimated probabilities towards the observed event fraction, such as ridge regression. Leave-one-out cross-validation also provided pessimistic estimates of the discrimination slope but nearly unbiased estimates of the Brier score.

Conclusions: We recommend to use leave-pair-out cross-validation, fivefold cross-validation with repetitions, the enhanced or the .632+ bootstrap to estimate c-statistics, and leave-pair-out or fivefold cross-validation to estimate discrimination slopes.