Bayesian meta-analysis of diagnostic tests allowing for imperfect reference standards

MetaBayesDTA: codeless Bayesian meta-analysis of test accuracy, with or without a gold standard

BACKGROUND

The statistical models developed for meta-analysis of diagnostic test accuracy studies require specialised knowledge to implement. This is especially true since recent guidelines, such as those in Version 2 of the Cochrane Handbook of Systematic Reviews of Diagnostic Test Accuracy, advocate more sophisticated methods than previously. This paper describes a web-based application - MetaBayesDTA - that makes many advanced analysis methods in this area more accessible.

RESULTS

We created the app using R, the Shiny package and Stan. It allows for a broad array of analyses based on the bivariate model including extensions for subgroup analysis, meta-regression and comparative test accuracy evaluation. It also conducts analyses not assuming a perfect reference standard, including allowing for the use of different reference tests.

CONCLUSIONS

Due to its user-friendliness and broad array of features, MetaBayesDTA should appeal to researchers with varying levels of expertise. We anticipate that the application will encourage higher levels of uptake of more advanced methods, which ultimately should improve the quality of test accuracy reviews.

Meta-analysis of dichotomous and ordinal tests with an imperfect gold standard

Standard methods for the meta-analysis of medical tests, without assuming a gold standard, are limited to dichotomous data. Multivariate probit models are used to analyse correlated dichotomous data, and can be extended to model ordinal data. Within the context of an imperfect gold standard, they have previously been used for the analysis of dichotomous and ordinal test data from a single study, and for the meta-analysis of dichotomous tests. However, they have not previously been used for the meta-analysis of ordinal tests. In this article, we developed a Bayesian multivariate probit latent class model for the simultaneous meta-analysis of ordinal and dichotomous tests without assuming a gold standard, which also allows one to obtain summary estimates of joint test accuracy. We fitted the models using the software Stan, which uses a state-of-the-art Hamiltonian Monte Carlo algorithm, and we applied the models to a dataset in which studies evaluated the accuracy of tests, and test combinations, for deep vein thrombosis. We demonstrate the issues with dichotomising ordinal test accuracy data in the presence of an imperfect gold standard, before applying and comparing several variations of our proposed model which do not require the data to be dichotomised. The models proposed will allow researchers to more appropriately meta-analyse ordinal and dichotomous tests without a gold standard, potentially leading to less biased estimates of test accuracy. This may lead to a better understanding of which tests, and test combinations, should be used for any given medical condition.

Prognostic and diagnostic accuracy of intracoronary electrocardiogram recorded during percutaneous coronary intervention: a meta-analysis

OBJECTIVE

Intracoronary ECG (IC-ECG) recording has been shown to be sensitive and reliable for detecting myocardial viability and local myocardial ischaemia in some studies. But IC-ECG is neither widely used during percutaneous coronary intervention (PCI) nor recommended in guidelines. This up-to-date meta-analysis of published studies was conducted to evaluate the prognostic and diagnostic accuracy of IC-ECG recorded during PCI.

METHODS

Relevant studies were identified by searches of MEDLINE until 19 June 2021. Observational and diagnostic studies which reported the prognostic or diagnostic accuracy of IC-ECG were included. Data were extracted independently by two authors. Summary estimates of clinical outcomes were obtained using a random effects model. Summary diagnostic accuracy was obtained by using a Bayesian bivariate random effects model.

RESULTS

Of the 12 included studies, 7 studies reported the clinical outcomes (821 patients) and 6 studies reported the diagnostic accuracy (485 patients) of IC-ECG. The pooled ORs with 95% CIs of ST-segment elevation recorded by IC-ECG were 4.65 (1.69 to 12.77), 5.08 (1.10 to 23.44), 4.53 (0.79 to 25.90) and 1.83 (0.93 to 3.62) for major adverse cardiac events, myocardial infarction, cardiac death and revascularisation, respectively. The weighted mean difference were 6.49 (95% CIs 3.84 to 9.14) for ejection fraction when ST-segment resolution was recorded, and 0.86 (95% CIs -8.55 to 10.26) when ST-segment elevation was recorded. The pooled sensitivity and specificity of ST-segment elevation were 0.78 (95% credibility intervals 0.64 to 0.89) and 0.87 (95% credibility intervals 0.75 to 0.94), respectively.

CONCLUSIONS

These findings provide quantitative data supporting that IC-ECG had promising diagnostic ability for local myocardial injury, and could predict clinical outcomes.

Diagnosis of visceral and cutaneous leishmaniasis using loop-mediated isothermal amplification (LAMP) protocols: a systematic review and meta-analysis

Sensitive, reliable and fast diagnostic tools that are applicable in low-resource settings, at the point of care (PoC), are seen as crucial in the fight against visceral leishmaniasis (VL) and cutaneous leishmaniasis (CL). Addressing the need for a PoC test, several diagnostic tests, including serological and molecular methods, have been developed and evaluated in the past. One promising molecular method, already implemented for diagnosis of a range of diseases, is the loop-mediated isothermal amplification (LAMP) protocol. In this systematic review and meta-analysis, using a comprehensive search strategy, we focus on studies evaluating the performance of LAMP for the diagnosis of leishmaniasis in humans and other mammals such as dogs, compared with microscopy and/or any other molecular diagnostic method. A meta-analysis, pooling sensitivity and specificity rates and calculating areas under the curve (AUCs) in summary receiver operating characteristic (SROC) plots, was conducted on datasets extracted from studies, grouped by clinical condition and sample type. We found high sensitivity and specificity for LAMP when compared with microscopy and PCR using blood samples, with pooled estimate values of > 90% for all subgroups, corresponding to calculated AUC values > 0.96, except for LAMP compared to microscopy for diagnosis of CL. However, only a limited number of studies were truly comparable. Most of the observed heterogeneity is likely based on true differences between the studies rather than sampling error only. Due to simple readout methods and low laboratory equipment requirements for sample preparation compared to other molecular methods, LAMP is a promising candidate for a molecular (near-)PoC diagnostic method for VL and CL.

Revealing a Novel Antigen Repressor of Differentiation Kinase 2 for Diagnosis of Human Visceral Leishmaniasis in India

Visceral leishmaniasis (VL) is one of the major global health concerns due to its association with morbidity and mortality. All available diagnostic tools have been, until now, unable to provide a very specific and cost-effective mode of detection for VL globally. Therefore, the design of robust, specific, and commercially translatable diagnostic tests is urgently required. Currently, we are attempting to identify and explore the diagnostic potential of a novel parasite antigen. Repressor of differentiation kinase 2 (RDK2), a serine/threonine kinase, has a versatile role in parasite life cycle progression. However, its role as a diagnostic candidate for VL has not been investigated. Herein, we cloned and over-expressed LdRDK2 and studied the recombinant RDK2 for the diagnosis of human VL using serum and urine samples. In silico analysis predicted that RDK2 is conserved among Leishmania species with the least conservation in humans. RDK2 developed immune-reactive bands with antibodies present in VL patients' sera, and it demonstrated no cross-reactivity with sera from healthy controls and other diseases. Additionally, RDK2 antigen demonstrated a significant reactivity with IgG antibodies of VL patients' sera, with 78% sensitivity and 86.67% specificity as compared to healthy controls and other diseases. Furthermore, we evaluated its utility for non-invasive diagnosis of VL using patients' urine samples and found 93.8% sensitivity and 85.7% specificity. RDK2 was found to have better sensitivity and treatment response in patients' urine compared to serum samples, indicating its role as a promising point of care (POC) antigen. In a nutshell, we explored the role of RDK2 as a potential diagnostic marker for VL in both invasive and non-invasive modes as well as its utility as a promising POC antigen for treatment response cases.

Diagnostic Performance of SGA, PG-SGA and MUST for Malnutrition Assessment in Adult Cancer Patients: A Systematic Literature Review and Hierarchical Bayesian Meta-Analysis

Because cancer-associated malnutrition is a major health complication, timely nutritional screening is of utmost importance. The aim of this study was to conduct a systematic literature review and meta-analysis of three tools in order to identify the method with the best diagnostic performance. PubMed, EMBASE and Cochrane central register of controlled trials were searched for articles published from database inception to January 2021. Studies that assessed the diagnostic accuracy of the SGA, PG-SGA or MUST in adult cancer patients were included. In order to evaluate the quality of each included study, the QUADAS-2 tool was used after which a meta-analysis was conducted using the hierarchical bivariate model. This model accounts for both within and between study variability. 16 studies (18 datasets) were included to evaluate these tools. The overall sensitivity and specificity for SGA was 0.69 and 0.80, 0.95 and 0.81 for PG-SGA, along with 0.83 and 0.83 for MUST respectively. An assessment of the likelihood ratios showed that PG-SGA had the highest LR + and the lowest LR-, it therefore has the best diagnostic performance to confirm malnutrition in adult cancer patients.

Prevalence of human visceral leishmaniasis in Iran: A systematic review and meta-analysis

Leishmania infantum is the main cause of human visceral leishmaniasis (HVL; also known as kala-azar) in the Middle East and may be fatal if left untreated. This disease was first reported in 1949 in Iran. Despite marked improvements in hygiene and sanitation conditions, the disease is still endemic in some parts of Iran. It is difficult to determine the current prevalence of HVL in Iran due to the scarcity of comprehensive studies in this regard. In response to this gap, a systematic review and meta-analysis was conducted to gain better understanding of HVL epidemiology in the general population of Iran. English and Persian databases were searched for studies reporting the prevalence and risk factors associated with HVL in the Iranian people from January 1995 to December 2019. The reported data were selected according to inclusion and exclusion criteria. The pooled prevalence of HVL infection and its 95 % confidence intervals were calculated. Quality assessment, heterogeneity testing and publication bias assessment were also done. Literature search revealed 3634 studies, of which 35 studies met our eligibility criteria, with a total of 50,716 individuals. The meta-analysis was performed on 31 out of 35 included studies. The estimated pooled prevalence of HVL infection according to seropositivity was 2% (95 % CI: 1-2%) in the general population of Iran in which western and northern provinces had the lowest and the highest prevalence, 0.5 % (95 %CI, 0.2-0.7%) and 3% (95 %CI, 1-5 %), respectively. The seroprevalence of HVL among females (2%; 95 %CI, 1-2 %) was more than males (1%; 95 %CI, 1-2 %). The ≤10 and >10 years age groups had similar seroprevalence rates (1%, 95 %CI, 1-2 % versus 1%, 95 %CI, 0-1 %, respectively). There was no significant difference in terms of geographic area, age and sex. Of 31 studies included in the meta-analysis, the most common diagnostic test was the direct agglutination test (96.77 %). To the best of our knowledge, this is the first systematic review of the prevalence of HVL in Iran. The results showed a low seroprevalence of HVL infection. However, the lack of published reports of HVL in an area does not exclusively mean the absence of the disease or carrier. We therefore recommend further studies in this regard.

Evaluation of Cysteine Protease C of Leishmania donovani in Comparison with Glycoprotein 63 and Elongation Factor 1α for Diagnosis of Human Visceral Leishmaniasis and for Posttreatment Follow-Up Response

Visceral leishmaniasis (VL) is a threat in many developing countries. Much effort has been put to eliminating this disease, for which serodiagnosis remains the mainstay for VL control programs. New and improved antigens as diagnostic candidates are required, though, as the available antigens fail to demonstrate equal optimum performance in all areas of endemicity. Moreover, these diagnoses are dependent on invasive serum sampling. In the current study, we cloned and expressed Leishmania donovani cysteine protease C (CPC) and evaluated its diagnostic and test-of-cure possibilities by detecting the antibody levels in human serum and urine through ELISA and immunoblot assays. Two immunodominant antigens, recombinant glycoprotein 63 (GP63) and elongation factor 1α (EF1α), identified earlier by our group, were also assessed by employing human serum and urine samples. Of these three antigens in ELISAs, CPC demonstrated the highest sensitivities of 98.15% and 96% positive testing in serum and urine of VL patients, respectively. Moreover, CPC yielded 100% specificity with serum and urine of nonendemic healthy controls compared to GP63 and EF1α. Urine samples were found to be more specific than serum for distinguishing endemic healthy controls and other diseases by means of all three antigens. In all cases, CPC gave the most promising results. Unlike serum, urine tests demonstrated a significant decrease in antibody levels for CPC, GP63, and EF1α after 6 months of treatment. The diagnostic and test-of-cure performances of CPC in the immunoblot assay were found to be better than those of GP63 and EF1α. In conclusion, CPC, followed by GP63 and EF1α, may be utilized as candidates for diagnosis of VL and to assess treatment response.

Nutritional screening tools for adult cancer patients: A hierarchical Bayesian latent-class meta-analysis

BACKGROUND & AIMS

Cancer treatment requires attentiveness to its broader effect on the body. Cancer's effect on appetite, strength, and body composition is contained in the summary term malnutrition. The tools used to detect malnutrition are a critical part of effective cancer care. In clinical care, selection of any specific tool is random. The relative validity of these tools have not been systematically compared. Using hierarchical Bayesian latent-class meta-analysis methods, this report compares three tools used for adult cancer patients - the Mini Nutritional Assessment (MNA), the Nutritional Risk Screening 2002 (NRS-2002) and the Patient Generated Subjective Global Assessment (PG-SGA).

METHOD

Drawing from English and Chinese language databases, a broad pool of eligible studies were identified for further selection and assessment. Using the hierarchical summary receiver operating characteristic (HSROC) model, pooled sensitivity, specificity, and other measurements the accuracy of the three tools were compared.

RESULT

A total of 37 eligible studies involving the MNA, NRS-2002 and PG-SGA were included in this meta-analysis. The pooled sensitivity was 0.910 (95% CI: 0.763 to 0.970) for MNA, 0.747 (95% CI: 0.680 to 0.804) for NRS-2002, and 0.964 (95% CI: 0.913 to 0.986) for PG-SGA. The pooled specificity was 0.720 (95% CI: 0.623 to 0.800) for MNA, 0.854 (95% CI: 0.808 to 0.891) for NRS-2002, 0.905 (95% CI: 0.807 to 0.956) for PG-SGA, respectively. The back-calculated likelihood ratio (LR) showed that MNA had a low negative likelihood ratio (LR^-), NRS-2002 corresponded to a high positive likelihood ratio (LR⁺) and PG-SGA represented the best LR⁺ and LR^-.

CONCLUSIONS

While there is no standard approach to assessment of malnutrition, the PG-SGA has the best diagnostic performance with cancer patients. Further work is needed to refine the utility of these tools in larger clinical samples.

Which Are the Optimal Criteria for the Diagnosis of Allergic Bronchopulmonary Aspergillosis? A Latent Class Analysis

BACKGROUND

The ideal criteria for diagnosing allergic bronchopulmonary aspergillosis (ABPA) remain unknown because of the lack of a criterion standard. Latent class analysis using a probabilistic modeling technique can circumvent the need for a reference standard.

OBJECTIVE

To compare the diagnostic performance of various criteria used for evaluating ABPA.

METHODS

We prospectively enrolled consecutive cases of bronchial asthma and performed a series of investigations used for the diagnosis of ABPA. We used latent class analysis to analyze the performance of various existing and novel diagnostic criteria.

RESULTS

Of the 543 subjects (mean age, 37 years; 319 women), 338 (62.2%) and 205 (37.8%) were labeled as "mild-to-moderate" and "severe" asthma cases, respectively. The subjects with severe asthma had a longer duration of asthma and a higher number of exacerbations in the previous year. The prevalence of Aspergillus fumigatus sensitization was 41% and 30%, using the A fumigatus-specific IgE and skin test, respectively. The prevalence of ABPA was 16%, using both the Rosenberg-Patterson and the International Society for Human and Animal Mycology (ISHAM)-ABPA Working Group criteria. The ISHAM criteria were slightly more sensitive (89% vs 81%) and specific (99% vs 98%) than the Patterson criteria. We obtained optimal diagnostic performance by altering the existing ISHAM criteria (serum total IgE >500 international units/mL, excluding the skin test, and using computed tomography of thorax instead of chest radiograph).

CONCLUSIONS

The ISHAM-ABPA Working Group criteria were only marginally better than the Patterson criteria in diagnosing ABPA among patients with asthma younger than 66 years. The diagnostic performance however improved by modifying the prevailing ISHAM criteria, but with increased cost.

An epidemiological study of visceral leishmaniasis in North East Ethiopia using serological and leishmanin skin tests

Background

In Ethiopia, visceral leishmaniasis (VL) is caused by Leishmania donovani. The estimated country-wide incidence of VL in Ethiopia is 3700–7400 cases/year. The balance between anthroponotic and zoonotic transmission is still unknown even though most authors believe that visceral leishmaniasis in East Africa is anthroponotic. Asymptomatic leishmania infections occur more frequently than clinically apparent visceral leishmaniasis cases. The aim of this study was to determine the prevalence of asymptomatic VL infection and assess the degree of exposure among residents in Raya Azebo Woreda villages where cases of VL were recently reported.

Methods

A community based cross-sectional survey was conducted in 2013 between 1^st of May and 25^th of July. A total of 1099 individuals living in 314 households were included in the study. Socio-demographic and clinical data were collected from each of the participants and venous blood was also collected for the detection of antibodies to visceral leishmaniasis using Direct Agglutination Test. Leishmanin skin test was performed to detect the exposure to the parasite. Data was entered into excel and exported to SPSS version 17 for statistical analysis. Chi-square and the corresponding p-values were used to determine the statistical significance of the proportions/ratios obtained from the cross tabulated data. A p-value < 0.05 was considered statistically significant.

Result

A total of 1099 study subjects comprising 401 males and 698 females were included in the study. The overall positive leishmanian skin test and sero-prevalence rates respectively were 9.08% and 0.87%. The difference in LST positivity by age group and sero-prevalence by sex were statistically significant (P <0.01 and P<0.05 respectively). Out of the 9 sero-positive individuals, 7 had no history of travel to visceral leishmaniasis endemic areas out of Raya Azebo.

Conclusion

In general our results suggest occurrence of VL in the study area is, very low. Our survey also indicates that due to the low incidence of the disease, and lack of awareness, some patients remain under diagnosed.

Diagnostic accuracy of DAT-SPECT and MIBG scintigraphy for dementia with Lewy bodies: an updated systematic review and Bayesian latent class model meta-analysis

PURPOSE

Imperfect clinical reference standards can preclude accurately estimating the diagnostic accuracy of DAT-SPECT and MIBG myocardial scintigraphy for diagnosing DLB. To investigate the validity of unadjusted accuracy, we updated our previous meta-analysis.

METHODS

Literature search was updated to March 18, 2018. We also examined published systematic review reports. Two investigators extracted data and rated study validity using the QUADAS-2 tool. We performed a Bayesian latent class model meta-analysis accounting for imperfect reference standards.

RESULTS

We evaluated 27 studies including 2236 patients. With the exception of two DAT-SPECT studies that involved postmortem neuropathological verification, studies were susceptible to bias from imperfect reference standards. Compared with the unadjusted accuracy estimates, the adjusted sensitivity values were similar, whereas the adjusted specificity values were generally lower for detecting α-synuclein pathology in the brain. The adjusted summary sensitivity and specificity were 0.86 (95% credible interval [CrI], 0.76-0.95) and 0.81 (CrI, 0.70-0.92), and 0.93 (CrI, 0.74-1.00) and 0.75 (CI, 0.47-0.94) for visual and semi-quantitative assessments of DAT-SPECT, respectively; 0.92 (CrI, 0.81-0.99) and 0.80 (CrI, 0.67-0.93), and 0.87 (CrI, 0.74-0.98) and 0.80 (CrI, 0.69-0.93), for delayed- and early-phase scans of MIBG scintigraphy, respectively. When diagnosing the typical clinical syndrome, the adjusted accuracy values were similar to the unadjusted estimates. The adjusted sensitivity and specificity were 0.89 (CrI, 0.75-0.98) and 0.87 (CrI, 0.72-0.97), and 0.97 (CrI, 0.78-1.0) and 0.70 (CrI, 0.43-0.92) for visual and semi-quantitative assessments of DAT-SPECT, respectively; and 0.93 (CrI, 0.81-0.98) and 0.90 (CrI, 0.73-0.97), and 0.85 (CrI, 0.66-0.96) and 0.96 (95% CI, 0.83-1.0) for delayed- and early-phase scans of MIBG scintigraphy, respectively.

CONCLUSIONS

In our adjusted analyses, both imaging biomarkers had high diagnostic accuracy for detecting the hallmark pathology in the brain and for diagnosing the typical clinical syndrome.

Evaluating the accuracy and economic value of a new test in the absence of a perfect reference test

BACKGROUND

Streptococcus pneumoniae (SP) pneumonia is often treated empirically as diagnosis is challenging because of the lack of a perfect test. Using BinaxNOW-SP, a urinary antigen test, as an add-on to standard cultures may not only increase diagnostic yield but also increase costs.

OBJECTIVE

To estimate the sensitivity and specificity of BinaxNOW-SP and subsequently estimate the cost-effectiveness of adding BinaxNOW-SP to the diagnostic work-up.

DESIGN

We fit a Bayesian latent-class meta-analysis model to obtain estimates of BinaxNOW-SP accuracy that adjust for the imperfect accuracy of culture. Meta-analysis results were combined with information on prevalence of SP pneumonia to estimate the number of patients who are correctly classified under competing diagnostic strategies. Taking into consideration the cost of antibiotics, we determined the incremental cost of adding BinaxNOW-SP to the work-up per case correctly diagnosed.

RESULTS

The BinaxNOW-SP test had a pooled sensitivity of 0.74 (95% credible interval [CrI], 0.67-0.83) and a pooled specificity of 0.96 (95% CrI, 0.92-0.99). An overall increase in diagnostic accuracy of 6.2% due to the addition of BinaxNOW-SP corresponded to an incremental cost per case correctly classified of $582 Canadian dollars.

CONCLUSIONS

The methods we have described allow us to evaluate the accuracy and economic value of a new test in the absence of a perfect reference test using an evidence-based approach.

Bayesian bivariate meta-analysis of diagnostic test studies with interpretable priors

In a bivariate meta-analysis, the number of diagnostic studies involved is often very low so that frequentist methods may result in problems. Using Bayesian inference is particularly attractive as informative priors that add a small amount of information can stabilise the analysis without overwhelming the data. However, Bayesian analysis is often computationally demanding and the selection of the prior for the covariance matrix of the bivariate structure is crucial with little data. The integrated nested Laplace approximations method provides an efficient solution to the computational issues by avoiding any sampling, but the important question of priors remain. We explore the penalised complexity (PC) prior framework for specifying informative priors for the variance parameters and the correlation parameter. PC priors facilitate model interpretation and hyperparameter specification as expert knowledge can be incorporated intuitively. We conduct a simulation study to compare the properties and behaviour of differently defined PC priors to currently used priors in the field. The simulation study shows that the PC prior seems beneficial for the variance parameters. The use of PC priors for the correlation parameter results in more precise estimates when specified in a sensible neighbourhood around the truth. To investigate the usage of PC priors in practice, we reanalyse a meta-analysis using the telomerase marker for the diagnosis of bladder cancer and compare the results with those obtained by other commonly used modelling approaches. Copyright © 2017 John Wiley & Sons, Ltd.

A joint latent class analysis for adjusting survival bias with application to a trauma transfusion study

There is no clear classification rule to rapidly identify trauma patients who are severely hemorrhaging and may need substantial blood transfusions. Massive transfusion (MT), defined as the transfusion of at least 10 units of red blood cells within 24 h of hospital admission, has served as a conventional surrogate that has been used to develop early predictive algorithms and establish criteria for ordering an MT protocol from the blood bank. However, the conventional MT rule is a poor proxy, because it is likely to misclassify many severely hemorrhaging trauma patients as they could die before receiving the 10th red blood cells transfusion. In this article, we propose to use a latent class model to obtain a more accurate and complete metric in the presence of early death. Our new approach incorporates baseline patient information from the time of hospital admission, by combining respective models for survival time and usage of blood products transfused within the framework of latent class analysis. To account for statistical challenges, caused by induced dependent censoring inherent in 24-h sums of transfusions, we propose to estimate an improved standard via a pseudo-likelihood function using an expectation-maximization algorithm with the inverse weighting principle. We evaluated the performance of our new standard in simulation studies and compared with the conventional MT definition using actual patient data from the Prospective Observational Multicenter Major Trauma Transfusion study. Copyright © 2015 John Wiley & Sons, Ltd.

Developments in Diagnosis of Visceral Leishmaniasis in the Elimination Era

Visceral leishmaniasis (VL) is the most devastating parasitic infection worldwide causing high morbidity and mortality. Clinical presentation of VL ranges from asymptomatic or subclinical infection to severe and complicated symptomatic disease. A major challenge in the clinical management of VL is the weakness of health systems in disease endemic regions. People affected by VL mostly present to primary health care centers (PHCs), often late in their therapeutic itinerary. PHC physicians face a major challenge: they do not deal with a single disease issue but with patients presenting with complaints pointing to several diagnostic possibilities. Risk exists when some patients having less clinical manifestations are misdiagnosed. Therefore, field based accurate, sensitive, and cost effective rapid diagnostic tools that can detect disease in its mildest form are essential for effective control and reaching the goal of VL elimination. In this review, we discuss the current status and challenges of various diagnostic tools for the diagnosis of VL and assess their application in resource poor settings.

A joint latent class model for classifying severely hemorrhaging trauma patients

BACKGROUND

In trauma research, "massive transfusion" (MT), historically defined as receiving ≥10 units of red blood cells (RBCs) within 24 h of admission, has been routinely used as a "gold standard" for quantifying bleeding severity. Due to early in-hospital mortality, however, MT is subject to survivor bias and thus a poorly defined criterion to classify bleeding trauma patients.

METHODS

Using the data from a retrospective trauma transfusion study, we applied a latent-class (LC) mixture model to identify severely hemorrhaging (SH) patients. Based on the joint distribution of cumulative units of RBCs and binary survival outcome at 24 h of admission, we applied an expectation-maximization (EM) algorithm to obtain model parameters. Estimated posterior probabilities were used for patients' classification and compared with the MT rule. To evaluate predictive performance of the LC-based classification, we examined the role of six clinical variables as predictors using two separate logistic regression models.

RESULTS

Out of 471 trauma patients, 211 (45 %) were MT, while our latent SH classifier identified only 127 (27 %) of patients as SH. The agreement between the two classification methods was 73 %. A non-ignorable portion of patients (17 out of 68, 25 %) who died within 24 h were not classified as MT but the SH group included 62 patients (91 %) who died during the same period. Our comparison of the predictive models based on MT and SH revealed significant differences between the coefficients of potential predictors of patients who may be in need of activation of the massive transfusion protocol.

CONCLUSIONS

The traditional MT classification does not adequately reflect transfusion practices and outcomes during the trauma reception and initial resuscitation phase. Although we have demonstrated that joint latent class modeling could be used to correct for potential bias caused by misclassification of severely bleeding patients, improvement in this approach could be made in the presence of time to event data from prospective studies.

A general framework for comparative Bayesian meta-analysis of diagnostic studies

BACKGROUND

Selecting the most effective diagnostic method is essential for patient management and public health interventions. This requires evidence of the relative performance of alternative tests or diagnostic algorithms. Consequently, there is a need for diagnostic test accuracy meta-analyses allowing the comparison of the accuracy of two or more competing tests. The meta-analyses are however complicated by the paucity of studies that directly compare the performance of diagnostic tests. A second complication is that the diagnostic accuracy of the tests is usually determined through the comparison of the index test results with those of a reference standard. These reference standards are presumed to be perfect, i.e. allowing the classification of diseased and non-diseased subjects without error. In practice, this assumption is however rarely valid and most reference standards show false positive or false negative results. When an imperfect reference standard is used, the estimated accuracy of the tests of interest may be biased, as well as the comparisons between these tests.

METHODS

We propose a model that allows for the comparison of the accuracy of two diagnostic tests using direct (head-to-head) comparisons as well as indirect comparisons through a third test. In addition, the model allows and corrects for imperfect reference tests. The model is inspired by mixed-treatment comparison meta-analyses that have been developed for the meta-analysis of randomized controlled trials. As the model is estimated using Bayesian methods, it can incorporate prior knowledge on the diagnostic accuracy of the reference tests used.

RESULTS

We show the bias that can result from using inappropriate methods in the meta-analysis of diagnostic tests and how our method provides more correct estimates of the difference in diagnostic accuracy between two tests. As an illustration, we apply this model to a dataset on visceral leishmaniasis diagnostic tests, comparing the accuracy of the RK39 dipstick with that of the direct agglutination test.

CONCLUSIONS

Our proposed meta-analytic model can improve the comparison of the diagnostic accuracy of competing tests in a systematic review. This is however only true if the studies and especially information on the reference tests used are sufficiently detailed. More specifically, the type and exact procedures used as reference tests are needed, including any cut-offs used and the number of subjects excluded from full reference test assessment. If this information is lacking, it may be better to limit the meta-analysis to direct comparisons.

Bayesian imperfect information analysis for clinical recurrent data

In medical research, clinical practice must often be undertaken with imperfect information from limited resources. This study applied Bayesian imperfect information-value analysis to realistic situations to produce likelihood functions and posterior distributions, to a clinical decision-making problem for recurrent events. In this study, three kinds of failure models are considered, and our methods illustrated with an analysis of imperfect information from a trial of immunotherapy in the treatment of chronic granulomatous disease. In addition, we present evidence toward a better understanding of the differing behaviors along with concomitant variables. Based on the results of simulations, the imperfect information value of the concomitant variables was evaluated and different realistic situations were compared to see which could yield more accurate results for medical decision-making.

Rapid tests for the diagnosis of visceral leishmaniasis in patients with suspected disease

BACKGROUND

The diagnosis of visceral leishmaniasis (VL) in patients with fever and a large spleen relies on showing Leishmania parasites in tissue samples and on serological tests. Parasitological techniques are invasive, require sophisticated laboratories, consume time, or lack accuracy. Recently, rapid diagnostic tests that are easy to perform have become available.

OBJECTIVES

To determine the diagnostic accuracy of rapid tests for diagnosing VL in patients with suspected disease presenting at health services in endemic areas.

SEARCH METHODS

We searched MEDLINE, EMBASE, LILACS, CIDG SR, CENTRAL, SCI-expanded, Medion, Arif, CCT, and the WHO trials register on 3 December 2013, without applying language or date limits.

SELECTION CRITERIA

This review includes original, phase III, diagnostic accuracy studies of rapid tests in patients clinically suspected to have VL. As reference standards, we accepted: (1) direct smear or culture of spleen aspirate; (2) composite reference standard based on one or more of the following: parasitology, serology, or response to treatment; and (3) latent class analysis.

DATA COLLECTION AND ANALYSIS

Two review authors independently extracted data and assessed quality of included studies using the QUADAS-2 tool. Discrepancies were resolved by a third author. We carried out a meta-analysis to estimate sensitivity and specificity of rapid tests, using a bivariate normal model with a complementary log-log link function. We analysed each index test separately. As possible sources of heterogeneity, we explored: geographical area, commercial brand of index test, type of reference standard, disease prevalence, study size, and risk of bias (QUADAS-2). We also undertook a sensitivity analysis to assess the influence of imperfect reference standards.

MAIN RESULTS

Twenty-four studies containing information about five index tests (rK39 immunochromatographic test (ICT), KAtex latex agglutination test in urine, FAST agglutination test, rK26 ICT, and rKE16 ICT) recruiting 4271 participants (2605 with VL) were included. We carried out a meta-analysis for the rK39 ICT (including 18 studies; 3622 participants) and the latex agglutination test (six studies; 1374 participants). The results showed considerable heterogeneity. For the rK39 ICT, the overall sensitivity was 91.9% (95% confidence interval (95% CI) 84.8 to 96.5) and the specificity 92.4% (95% CI 85.6 to 96.8). The sensitivity was lower in East Africa (85.3%; 95% CI 74.5 to 93.2) than in the Indian subcontinent (97.0%; 95% CI 90.0 to 99.5). For the latex agglutination test, overall sensitivity was 63.6% (95% CI 40.9 to 85.6) and specificity 92.9% (95% CI 76.7 to 99.2).

AUTHORS' CONCLUSIONS

The rK39 ICT shows high sensitivity and specificity for the diagnosis of visceral leishmaniasis in patients with febrile splenomegaly and no previous history of the disease, but the sensitivity is notably lower in east Africa than in the Indian subcontinent. Other rapid tests lack accuracy, validation, or both.