Reference Interval Harmonization: Harnessing the Power of Big Data Analytics to Derive Common Reference Intervals across Populations and Testing Platforms.

BACKGROUND: Harmonization in laboratory medicine is essential for consistent and accurate clinical decision-making. There is significant and unwarranted variation in reference intervals (RIs) used by laboratories for assays with established analytical traceability. The Canadian Society of Clinical Chemists (CSCC) Working Group on Reference Interval Harmonization (hRI-WG) aims to establish harmonized RIs (hRIs) for laboratory tests and support implementation. METHODS: Harnessing the power of big data, laboratory results were collected across populations and testing platforms to derive common adult RIs for 16 biochemical markers. A novel comprehensive approach was established, including: (a) analysis of big data from community laboratories across Canada; (b) statistical evaluation of age, sex, and analytical differences; (c) derivation of hRIs using the refineR method; and (d) verification of proposed hRIs across 9 laboratories with different instrumentation using serum and plasma samples collected from healthy Canadian adults. RESULTS: Harmonized RIs were calculated for all assays using the refineR method, except free thyroxine. Derived hRIs met proposed verification criterion across 9 laboratories and 5 manufacturers for alkaline phosphatase, albumin (bromocresol green), chloride, lactate dehydrogenase, magnesium, phosphate, potassium (serum), and total protein (serum). Further investigation is needed for some analytes due to failure to meet verification criteria in one or more laboratories (albumin [bromocresol purple], calcium, total carbon dioxide, total bilirubin, and sodium) or concern regarding excessively wide hRIs (alanine aminotransferase, creatinine, and thyroid stimulating hormone). CONCLUSIONS: We report a novel data-driven approach for RI harmonization. Findings support feasibility of RI harmonization for several analytes; however, some presented challenges, highlighting limitations that need to be considered in harmonization and big data analytics.

Detection of Large Genomic RNA via DNAzyme-Mediated RNA Cleavage and Rolling Circle Amplification: SARS-CoV-2 as a Model.

A new method for the detection of genomic RNA combines RNA cleavage by the 10-23 DNAzyme and use of the cleavage fragments as primers to initiate rolling circle amplification (RCA). 230 different 10-23 DNAzyme variants were screened to identify those that target accessible RNA sites within the highly structured RNA transcripts of SARS-CoV-2. A total of 28 DNAzymes were identified with >20 % cleavage, 5 with >40 % cleavage and one with >60 % in 10 min. The cleavage fragments from these reactions were then screened for coupling to an RCA reaction, leading to the identification of several cleavage fragments that could efficiently initiate RCA. Using a newly developed quasi-exponential RCA method with a detection limit of 500 aM of RNA, 14 RT-PCR positive and 15 RT-PCR negative patient saliva samples were evaluated for SARS-CoV-2 genomic RNA, achieving a clinical sensitivity of 86 % and specificity of 100 % for detection of the virus in <2.5 h.

Diverse high-affinity DNA aptamers for wild-type and B.1.1.7 SARS-CoV-2 spike proteins from a pre-structured DNA library.

We performed in vitro selection experiments to identify DNA aptamers for the S1 subunit of the SARS-CoV-2 spike protein (S1 protein). Using a pool of pre-structured random DNA sequences, we obtained over 100 candidate aptamers after 13 cycles of enrichment under progressively more stringent selection pressure. The top 10 sequences all exhibited strong binding to the S1 protein. Two aptamers, named MSA1 (Kd = 1.8 nM) and MSA5 (Kd = 2.7 nM), were assessed for binding to the heat-treated S1 protein, untreated S1 protein spiked into 50% human saliva and the trimeric spike protein of both the wildtype and the B.1.1.7 variant, demonstrating comparable affinities in all cases. MSA1 and MSA5 also recognized the pseudotyped lentivirus of SARS-CoV-2 with respective Kd values of 22.7 pM and 11.8 pM. Secondary structure prediction and sequence truncation experiments revealed that both MSA1 and MSA5 adopted a hairpin structure, which was the motif pre-designed into the original library. A colorimetric sandwich assay was developed using MSA1 as both the recognition element and detection element, which was capable of detecting the pseudotyped lentivirus in 50% saliva with a limit of detection of 400 fM, confirming the potential of these aptamers as diagnostic tools for COVID-19 detection.

Three on Three: Universal and High-Affinity Molecular Recognition of the Symmetric Homotrimeric Spike Protein of SARS-CoV-2 with a Symmetric Homotrimeric Aptamer.

Our previously discovered monomeric aptamer for SARS-CoV-2 (MSA52) possesses a universal affinity for COVID-19 spike protein variants but is ultimately limited by its ability to bind only one subunit of the spike protein. The symmetrical shape of the homotrimeric SARS-CoV-2 spike protein presents the opportunity to create a matching homotrimeric molecular recognition element that is perfectly complementary to its structural scaffold, causing enhanced binding affinity. Here, we describe a branched homotrimeric aptamer with three-fold rotational symmetry, named TMSA52, that not only possesses excellent binding affinity but is also capable of binding several SARS-CoV-2 spike protein variants with picomolar affinity, as well as pseudotyped lentiviruses expressing SARS-CoV-2 spike protein variants with femtomolar affinity. Using Pd-Ir nanocubes as nanozymes in an enzyme-linked aptamer binding assay (ELABA), TMSA52 was capable of sensitively detecting diverse pseudotyped lentiviruses in pooled human saliva with a limit of detection as low as 6.3 × 103 copies/mL. The ELABA was also used to test 50 SARS-CoV-2-positive and 60 SARS-CoV-2-negative patient saliva samples, providing sensitivity and specificity values of 84.0 and 98.3%, respectively, thus highlighting the potential of TMSA52 for the development of future rapid tests.

High-Affinity Dimeric Aptamers Enable the Rapid Electrochemical Detection of Wild-Type and B.1.1.7 SARS-CoV-2 in Unprocessed Saliva.

We report a simple and rapid saliva-based SARS-CoV-2 antigen test that utilizes a newly developed dimeric DNA aptamer, denoted as DSA1N5, that specifically recognizes the spike proteins of the wildtype virus and its Alpha and Delta variants with dissociation constants of 120, 290 and 480 pM, respectively, and binds pseudotyped lentiviruses expressing the wildtype and alpha trimeric spike proteins with affinity constants of 2.1 pM and 2.3 pM, respectively. To develop a highly sensitive test, DSA1N5 was immobilized onto gold electrodes to produce an electrochemical impedance sensor, which was capable of detecting 1000 viral particles per mL in 1:1 diluted saliva in under 10 min without any further sample processing. Evaluation of 36 positive and 37 negative patient saliva samples produced a clinical sensitivity of 80.5 % and specificity of 100 % and the sensor could detect the wildtype virus as well as the Alpha and Delta variants in the patient samples, which is the first reported rapid test that can detect any emerging variant of SARS-CoV-2.

The relation between DNA methylation patterns and serum cytokine levels in community-dwelling adults: a preliminary study.

BACKGROUND: The levels of circulating cytokines fluctuate with age, acute illness, and chronic disease, and are predictive of mortality; this is also true for patterns of DNA (CpG) methylation. Given that immune cells are particularly sensitive to changes in the concentration of cytokines in their microenvironment, we hypothesized that serum levels of TNF, IL-6, IL-8 and IL-10 would correlate with genome-wide alterations in the DNA methylation levels of blood leukocytes. To test this, we evaluated community-dwelling adults (n = 14; 48-78 years old) recruited to a pilot study for the Canadian Longitudinal Study on Aging (CLSA), examining DNA methylation patterns in peripheral blood mononuclear cells using the Illumina HumanMethylation 450 K BeadChip. RESULTS: We show that, apart from age, serum IL-10 levels exhibited the most substantial association to DNA methylation patterns, followed by TNF, IL-6 and IL-8. Furthermore, while the levels of these cytokines were higher in elderly adults, no associations with epigenetic accelerated aging, derived using the epigenetic clock, were observed. CONCLUSIONS: As a preliminary study with a small sample size, the conclusions drawn from this work must be viewed with caution; however, our observations are encouraging and certainly warrant more suitably powered studies of this relationship.

Statistical methods used in the calculation of geriatric reference intervals: a systematic review.

BACKGROUND: Geriatric reference intervals (RIs) are not commonly available and are rarely used. It is difficult to select a reference population from a cohort with a high degree of morbidity. Also important are the statistical approaches used to determine health-associated reference values. It is the aim of this study to examine the statistical methods used in the calculation of geriatric RIs. METHODS: A search was conducted on EMBASE and Medline for articles between January 1989 and January 2014. Studies were selected if they: 1) were English primary articles; 2) performed a clinical chemistry test on a blood fraction; 3) had a population sub-group consisting of individuals ≥65 years of age; and 4) calculated a RI for the subgroup ≥65 years of age. RESULTS: There were 64 articles identified, of which 78.1% described the RI calculation method used. RI calculation was performed by non-parametric (21.9%), parametric (42.2%), robust (3.1%), or other (17.2%) methods. Outlier detection (SD, Grubb's test, Tukey's fence, Dixon) was infrequently used and although most studies performed partitioning, only 57.8% tested the statistical significance of the partitions. Few studies (17.2%) reported confidence intervals for the RI estimates. Overall, only 14.1% of studies provided RI estimates which followed the CLSI guideline EP28-A3c. CONCLUSIONS: Statistical methods for RI calculation and partitioning varied considerably between studies and many failed to provide adequate descriptions of these methods. Challenges in analyses arose from insufficient sample sizes and heterogeneity in the elderly population. Geriatric RIs, although present in the literature, may not be properly calculated and should be carefully considered before applying them for clinical care.

BNP and NT-proBNP as prognostic markers in persons with acute decompensated heart failure: a systematic review.

A systematic review was undertaken to examine the evidence for B-type natriuretic peptides (BNP and NT-proBNP) as independent predictors of mortality, morbidity, or combined mortality and morbidity outcomes in persons with acute decompensated heart failure (ADHF). Electronic databases (Medline(®), Embase™, AMED, Cochrane Central Register of Controlled Trials, Cochrane Database of Systematic Reviews, and CINAHL) were searched from 1989 to June 2012. Reference lists of included articles, systematic reviews, and the gray literature were also searched. English language studies were eligible if they included subjects with ADHF and measured BNP/NT-proBNP using FDA approved assays. Standardized forms were used to select studies, extract data, and assess risk of bias. Seventy-nine studies, ranging over followup intervals from 14 days to 7 years, evaluating levels of BNP (n = 38), NT-proBNP (n = 35), or both (n = 6) were eligible. The majority of studies predicted mortality outcomes for admission BNP/NT-proBNP levels, with fewer studies evaluating serial, change from admission, or discharge levels. In general, higher levels of admission BNP or NT-proBNP predicted greater risk for all outcomes. Decreased levels post-admission predicted decreased risk. Overall, these studies were rated as having moderate risk of bias. This systematic review shows that BNP and NT-proBNP are independent predictors of mortality (all-cause and cardiovascular) in ADHF despite different cutpoints, time intervals, and prognostic models. Findings for morbidity and composite outcomes were less frequently evaluated and showed inconsistency. Further research is required to assess cutpoints for admission, serial measurements, change following admission, and discharge levels to assist clinical decision-making.

Incremental predictive value of natriuretic peptides for prognosis in the chronic stable heart failure population: a systematic review.

The aim of this study was to determine whether measurement of natriuretic peptides independently adds incremental predictive value for mortality and morbidity in patients with chronic stable heart failure (CSHF). We electronically searched Medline®, Embase™, AMED, Cochrane Central Register of Controlled Trials, Cochrane Database of Systematic Reviews, and CINAHL from 1989 to June 2012. We also searched reference lists of included articles, systematic reviews, and the gray literature. Studies were screened for eligibility criteria and assessed for methodological quality. Data were extracted on study design, population demographics, assay cutpoints, prognostic risk prediction model covariates, statistical methods, outcomes, and results. One hundred and eighty-three studies were identified as prognostic in the systematic review. From these, 15 studies (all NT-proBNP) considered incremental predictive value in CSHF subjects. Follow-up varied from 12 to 37 months. All studies presented at least one estimate of incremental predictive value of NT-proBNP relative to the base prognostic model. Using discrimination or likelihood statistics, these studies consistently showed that NT-proBNP increased model performance. Three studies used re-classification and model validation computations to establish incremental predictive value; these studies showed less consistency with respect to added value. Although there were differences in the base risk prediction models, assay cutpoints, and lengths of follow-up, there was consistency in NT-proBNP adding incremental predictive value for prognostic models in chronic stable CSHF patients. The limitations in the literature suggest that studies designed to evaluate prognostic models should be undertaken to evaluate the incremental value of natriuretic peptide as a predictor of mortality and morbidity in CSHF.

B-type natriuretic peptide-guided therapy: a systematic review.

BNP/NT-proBNP measurement has not gained widespread use for the management of patients with heart failure (HF) despite several randomized controlled trials. A systematic review addressing the question of whether patients with HF benefit from BNP-assisted therapy or intensified therapy compared with usual care was undertaken. Relevant randomized controlled trial (RCTs) were selected by searching Medline, Embase, AMED, Cochrane Central Register of Controlled Trials, Cochrane Database of Systematic Reviews, and CINAHL for English-language articles published from 1980 to 2012. Selected studies required patients to be treated for chronic HF with medical therapy based on BNP/NT-proBNP or usual care. There were no restrictions except that BNP/NT-proBNP measurement had to be done by an FDA approved method. Nine RCTs were identified with 2,104 patients with study duration that ranged from 3 to 18 months. Overall, there was a wide variation in study design and how parameters were reported including patient selection, baseline characteristics, therapy goals, BNP/NT-proBNP cutpoint, and outcome types. Meta-analysis was not appropriate given this study heterogeneity. The strength of evidence for the outcome of mortality, reported in seven studies, was found to be low due to inconsistency and imprecision. This systematic review showed that the evidence is of low quality and insufficient to support the use of BNP/NT-proBNP to guide HF therapy. Further trials with improved design are needed.

Incremental value of natriuretic peptide measurement in acute decompensated heart failure (ADHF): a systematic review.

The aim of this systematic review was to determine whether B-type natriuretic peptide (BNP) and N-terminal proBNP (NT-proBNP) independently add incremental value for predicting mortality and morbidity in patients with acute decompensated heart failure (ADHF). Medline(®), Embase™, AMED, Cochrane Central Register of Controlled Trials, Cochrane Database of Systematic Reviews, and CINAHL were searched from 1989 to June 2012. We also searched reference lists of included articles, systematic reviews, and the gray literature. Studies were screened for eligibility criteria and assessed for risk of bias. Data were extracted on study design, population demographics, assay cutpoints, prognostic risk prediction model covariates, statistical methods, outcomes, and results. From 183 citations, only seven studies (5 BNP and 2 NT-proBNP) considered incremental value in ADHF subjects admitted to acute care centers. Admission assay levels and length of follow-up varied for BNP studies (31 days to 12 months) and for NT-proBNP studies (25-82 months). All studies presented at least one estimate of incremental value of BNP/NT-proBNP relative to the base prognostic model. Using discrimination or likelihood statistics, these studies consistently showed that BNP or NT-proBNP increased model performance. Three studies used reclassification and model validation computations to establish incremental value; these studies showed less consistency with respect to added value. In conclusion, the literature assessing incremental value of BNP/NT-proBNP in ADHF populations is limited to seven studies evaluating only mortality outcomes and at moderate risk of bias. Although there were differences in the base risk prediction models, assay cutpoints, and lengths of follow-up, there was consistency in BNP/NT-proBNP adding incremental value in prediction models in ADHF patients.

Prediction of clinical outcomes using B-type natriuretic peptides in the general population: a systematic review.

The use of B-type natriuretic peptides to predict outcomes in general populations has been investigated in a number of primary studies. A previous systematic review considering natriuretic peptides in cardiovascular disease included a subgroup of general population studies, which suggested an association with a number of clinical outcomes. We electronically searched Medline, Embase, AMED, Cochrane Central Register of Controlled Trials, Cochrane Database of Systematic Reviews, and CINAHL for English-language articles published between 1989 and mid-2012. We utilized trained reviewers and standardized forms to screen articles for inclusion and extract data from included articles. All included studies (n = 7) were summarized in narrative and tabular form. A general population was defined as one that was randomly selected from a community setting where no specific inclusion or exclusion criteria were specified. The seven included studies all used FDA approved assays for NT-proBNP. The range of clinical outcomes and heterogeneity did not allow for meta-analysis. The hazard ratios for predicting outcomes in the included studies ranged from 1.0 to 4.1 (all p values <0.05). The discrimination statistics reported in four studies all demonstrated statistically significant improvements in predicting outcomes. NT-proBNP is associated with heart failure, all-cause and cardiovascular mortality, and other combined cardiovascular events in a general unselected population. The discrimination statistics suggest modest improvements in risk stratification. No prospective studies exist to demonstrate the clinical utility of using B-type natriuretic peptides to predict clinical outcomes in a general population.

Use of BNP and NT-proBNP for the diagnosis of heart failure in the emergency department: a systematic review of the evidence.

Our purpose was to determine the test performance characteristics of BNP and NT-proBNP in the diagnosis of heart failure for patients presenting to an emergency department or urgent care center. We searched Medline, Embase, AMED, Cochrane, Cochrane Database of Systematic Reviews, and CINAHL for English-language articles published between 1989 and June 2012. Studies were limited to those using FDA-approved assays. We examined test performance at three pre-specified cutpoints (manufacturers' suggested, researchers' optimal, and lowest) and considered the effect of age, gender, ethnicity and renal function. We used the QUADAS-2 tool to examine risk of bias and applicability, and the AHRQ Methods Guide to assess the strength of evidence. Seventy-six articles met our inclusion criteria, 37 examined BNP, 25 examined NT-proBNP, and 14 examined both. Pooled sensitivity and specificity for BNP at the three pre-specified cutpoints were 95, 91, and 95 % (sensitivity) and 55, 80, and 67 % (specificity), respectively. For NT-proBNP, sensitivity and specificity at the same cutpoints were 91, 90, and 96 % (sensitivity) and 67, 74, and 55 % (specificity). Both BNP and NT-proBNP perform well to rule out, but less well to rule in, the diagnosis of heart failure among persons presenting to emergency departments or urgent care centers. Both BNP and NT-proBNP levels are positively associated with age and negatively associated with renal function. However, the effect of these factors with respect to selecting optimal cutpoints is unclear. For BNP, 100 pg/mL appears to be a consensus cutpoint. No clear consensus has emerged for NT-proBNP, but the age-adjusted cutpoints of 450 pg/mL for <50 years, 900 pg/mL for 50-75 years and 1,800 pg/mL for >75 years appear promising and merit greater scrutiny and validation.

Performance of BNP and NT-proBNP for diagnosis of heart failure in primary care patients: a systematic review.

National and international guidelines have been published recommending the use of natriuretic peptides as an aid to the diagnosis of heart failure (HF) in acute settings; however, few specific recommendations exist for governing the use of these peptides in primary care populations. To summarize the available data relevant to the diagnosis of HF in primary care patient population, we systematically reviewed the literature to identify original articles that investigated the diagnostic accuracy of B-type natriuretic peptide (BNP) and N-terminal proBNP (NT-proBNP) in primary care settings. The search yielded 25,864 articles in total: 12 investigating BNP and 20 investigating NT-proBNP were relevant to our objective and included in the review. QUADAS-2 and GRADE were used to assess the quality of the included articles. Diagnostic data were pooled based on three cutpoints: lowest and optimal, as chosen by study authors, and manufacturers' suggested. The effect of various determinants (e.g., age, gender, BMI, and renal function) on diagnostic performance was also investigated. Pooled sensitivity and specificity of BNP and NT-proBNP using the lowest [0.85 (sensitivity) and 0.54 (specificity)], optimal (0.80 and 0.61), and manufacturers' (0.74 and 0.67) cutpoints showed good performance for diagnosing HF. Similar performance was seen for NT-proBNP: lowest (0.90 and 0.50), optimal (0.86 and 0.58), and manufacturers' (0.82 and 0.58) cutpoints. Overall, we rated the strength of evidence as high because further studies will be unlikely to change the estimates diagnostic performance.

A systematic review of BNP and NT-proBNP in the management of heart failure: overview and methods.

B-type natriuretic peptide (BNP) and N-terminal proBNP (NT-proBNP) levels are increased in persons with heart failure (HF); low levels of these peptides rule out HF. We systematically reviewed the literature to assess the use of BNP and NT-proBNP in the diagnosis, prognosis, and treatment for HF. We also examined the biological variation of these peptides in persons with and without HF. We searched Medline, Embase, AMED, Cochrane Central Register of Controlled Trials, Cochrane Database of Systematic Reviews, and CINAHL for English-language studies published between January 1989 and June 2012. Supplemental searches involved the gray literature and the reference lists of included studies. Trained reviewers used standardized forms to screen articles for inclusion in the review and to extract data from included papers. We examined the risk of bias with QUADAS-2 for diagnosis studies, the Hayden criteria for prognosis studies, and the Jadad scale for treatment studies. We assessed the strength of evidence in four domains (risk of bias, consistency, directness, and precision) for the diagnosis and treatment studies. Results were reported as narrative syntheses. Additional meta-analyses were conducted for the diagnosis studies. Three hundred ten articles passed through screening and were included in the review. One hundred four articles applied to diagnostic accuracy, 190 papers pertained to prognosis, and nine articles addressed BNP- or NT-proBNP-guided treatment. Each individual paper in this series reports, summarizes, and discusses the evidence regarding diagnosis, prognosis, or treatment.

BNP and NT-proBNP as prognostic markers in persons with chronic stable heart failure.

Prognosis permits clinicians to separate persons with heart failure (HF) into subgroups based on likely health outcomes. Treatment is partly guided by these likely outcomes. This systematic review explores whether brain natriuretic peptide (BNP) and N-terminal proBNP (NT-proBNP) are independent predictors of prognosis in persons with chronic stable HF. We electronically searched Medline, Embase, AMED, Cochrane Central Register of Controlled Trials, Cochrane Database of Systematic Reviews, and CINAHL for English-language articles published between 1989 and mid-2012. We utilized trained reviewers and standardized forms to screen articles for inclusion and extracted data from included articles. All included studies were summarized in narrative and tabular form. We used the Hayden criteria to assess the risk of bias. Sixteen BNP publications and 88 NT-proBNP publications were included in the systematic review. BNP was positively associated with all-cause and HF mortality. NT-proBNP was positively associated with all-cause and cardiovascular mortality. BNP and NT-proBNP levels are useful for estimating prognosis in persons with chronic stable HF. Further research is required to establish optimal cutpoints and to assess whether prognostic effects differ by age, sex, or time period.

Hematocrit correction does not improve glucose monitor accuracy in the assessment of neonatal hypoglycemia.

BACKGROUND: The lack of accuracy of point of care (POC) glucose monitors has limited their use in the diagnosis of neonatal hypoglycemia. Hematocrit plays an important role in explaining discordant results. The objective of this study was to to assess the effect of hematocrit on the diagnostic performance of Abbott Precision Xceed Pro (PXP) and Nova StatStrip (StatStrip) monitors in neonates. METHODS: All blood samples ordered for laboratory glucose measurement were analyzed using the PXP and StatStrip and compared with the laboratory analyzer (ABL 800 Blood Gas analyzer [ABL]). Acceptable error targets were ±15% for glucose monitoring and ±5% for diagnosis. RESULTS: A total of 307 samples from 176 neonates were analyzed. Overall, 90% of StatStrip and 75% of PXP values met the 15% error limit and 45% of StatStrip and 32% of PXP values met the 5% error limit. At glucose concentrations ≤4 mmol/L, 83% of StatStrip and 79% of PXP values met the 15% error limit, while 37% of StatStrip and 38% of PXP values met the 5% error limit. Hematocrit explained 7.4% of the difference between the PXP and ABL whereas it accounted for only 0.09% of the difference between the StatStrip and ABL. The ROC analysis showed the screening cut point with the best performance for identifying neonatal hypoglycemia was 3.2 mmol/L for StatStrip and 3.3 mmol/L for PXP. CONCLUSIONS: Despite a negligible hematocrit effect for the StatStrip, it did not achieve recommended error limits. The StatStrip and PXP glucose monitors remain suitable only for neonatal hypoglycemia screening with confirmation required from a laboratory analyzer.

Performance of saliva compared with nasopharyngeal swab for diagnosis of COVID-19 by NAAT in cross-sectional studies: Systematic review and meta-analysis.

Nucleic acid amplification testing (NAAT) is the preferred method to diagnose coronavirus disease 2019 (COVID-19). Saliva has been suggested as an alternative to nasopharyngeal swabs (NPS), but previous systematic reviews were limited by the number and types of studies available. The objective of this systematic review and meta-analysis was to assess the diagnostic performance of saliva compared with NPS for COVID-19. We searched Ovid MEDLINE, Embase, Cochrane, and Scopus databases up to 24 April 2021 for studies that directly compared paired NPS and saliva specimens taken at the time of diagnosis. Meta-analysis was performed using an exact binomial rendition of the bivariate mixed-effects regression model. Risk of bias was assessed using the QUADAS-2 tool. Of 2683 records, we included 23 studies with 25 cohorts, comprising 11,582 paired specimens. A wide variety of NAAT assays and collection methods were used. Meta-analysis gave a pooled sensitivity of 87 % (95 % CI = 83-90 %) and specificity of 99 % (95 % CI = 98-99 %). Subgroup analyses showed the highest sensitivity when the suspected individual is tested in an outpatient setting and is symptomatic. Our results support the use of saliva NAAT as an alternative to NPS NAAT for the diagnosis of COVID-19.

Chapter 9: options for summarizing medical test performance in the absence of a "gold standard".

The classical paradigm for evaluating test performance compares the results of an index test with a reference test. When the reference test does not mirror the "truth" adequately well (e.g. is an "imperfect" reference standard), the typical ("naïve") estimates of sensitivity and specificity are biased. One has at least four options when performing a systematic review of test performance when the reference standard is "imperfect": (a) to forgo the classical paradigm and assess the index test's ability to predict patient relevant outcomes instead of test accuracy (i.e., treat the index test as a predictive instrument); (b) to assess whether the results of the two tests (index and reference) agree or disagree (i.e., treat them as two alternative measurement methods); (c) to calculate "naïve" estimates of the index test's sensitivity and specificity from each study included in the review and discuss in which direction they are biased; (d) mathematically adjust the "naïve" estimates of sensitivity and specificity of the index test to account for the imperfect reference standard. We discuss these options and illustrate some of them through examples.

Chapter 8: meta-analysis of test performance when there is a "gold standard".

Synthesizing information on test performance metrics such as sensitivity, specificity, predictive values and likelihood ratios is often an important part of a systematic review of a medical test. Because many metrics of test performance are of interest, the meta-analysis of medical tests is more complex than the meta-analysis of interventions or associations. Sometimes, a helpful way to summarize medical test studies is to provide a "summary point", a summary sensitivity and a summary specificity. Other times, when the sensitivity or specificity estimates vary widely or when the test threshold varies, it is more helpful to synthesize data using a "summary line" that describes how the average sensitivity changes with the average specificity. Choosing the most helpful summary is subjective, and in some cases both summaries provide meaningful and complementary information. Because sensitivity and specificity are not independent across studies, the meta-analysis of medical tests is fundamentaly a multivariate problem, and should be addressed with multivariate methods. More complex analyses are needed if studies report results at multiple thresholds for positive tests. At the same time, quantitative analyses are used to explore and explain any observed dissimilarity (heterogeneity) in the results of the examined studies. This can be performed in the context of proper (multivariate) meta-regressions.