Objective to review the evidence from studies relating SARS-CoV-2 culture with the results of reverse transcriptase polymerase chain reaction (RT-PCR) and other variables which may influence the interpretation of the test, such as time from symptom onset Methods We searched LitCovid, medRxiv, Google Scholar and the WHO Covid-19 database for Covid-19 to 10 September 2020. We included studies attempting to culture or observe SARS-CoV-2 in specimens with RT-PCR positivity. Studies were dual extracted and the data summarised narratively by specimen type. Where necessary we contacted corresponding authors of included papers for additional information. We assessed quality using a modified QUADAS 2 risk of bias tool. Results We included 29 studies reporting attempts at culturing, or observing tissue infection by, SARS-CoV-2 in sputum, nasopharyngeal or oropharyngeal, urine, stool, blood and environmental specimens. The quality of the studies was moderate with lack of standardised reporting. The data suggest a relationship between the time from onset of symptom to the timing of the specimen test, cycle threshold (Ct) and symptom severity. Twelve studies reported that Ct values were significantly lower and log copies higher in specimens producing live virus culture. Two studies reported the odds of live virus culture reduced by approximately 33% for every one unit increase in Ct. Six of eight studies reported detectable RNA for longer than 14 days but infectious potential declined after day 8 even among cases with ongoing high viral loads. Four studies reported viral culture from stool specimens. Conclusion Complete live viruses are necessary for transmission, not the fragments identified by PCR. Prospective routine testing of reference and culture specimens and their relationship to symptoms, signs and patient co-factors should be used to define the reliability of PCR for assessing infectious potential. Those with high cycle threshold are unlikely to have infectious potential.

We report the results of a review of the evidence from studies comparing SARS-CoV-2 culture with reverse transcriptase polymerase chain reaction (rt-PCR), as viral culture represents the best indicator of current infection and infectiousness of the isolate. We identified fourteen studies succeeding in culturing or observing tissue invasion by SARS-CoV in sputum, naso or oropharyngeal, urine, stool and environmental samples from patients diagnosed with Covid-19. The data are suggestive of a relation between the time from collection of a specimen to test, copy threshold, and symptom severity, but the quality of the studies was moderate with lack of standardised reporting and lack of testing of PCR against viral culture or infectivity in animals. This limits our current ability to quantify the relationship between viral load, cycle threshold and viable virus detection and ultimately the usefulness of PCR use for assessing infectiousness of patients. Prospective routine testing of reference and culture specimens are necessary for each country involved in the pandemic to establish the usefulness and reliability of PCR for Covid-19 and its relation to patients factors such as date of onset of symptoms and copy threshold, in order to help predict infectivity.

Objective To simulate possible changes in systematic review results if rapid review methods were used. Study Design and Setting We recalculated meta-analyses for binary primary outcomes in Cochrane systematic reviews, simulating rapid review methods. We simulated searching only PubMed, excluding older articles (5, 7, 10, 15, and 20 years before the search date), excluding smaller trials (<50, <100, and <200 participants), and using the largest trial only. We examined percentage changes in pooled odds ratios (ORs) (classed as no important change [<5%], small [<20%], moderate [<30%], or large [≥30%]), statistical significance, and biases observed using rapid methods. Results Two thousand five hundred and twelve systematic reviews (16,088 studies) were included. Rapid methods resulted in the loss of all data in 3.7–44.7% of meta-analyses. Searching only PubMed had the smallest risk of changed ORs (19% [477/2,512] were small changes or greater; 10% [260/2,512] were moderate or greater). Changes in ORs varied substantially with each rapid review method; 8.4–21.3% were small, 1.9–8.8% were moderate, and 4.7–34.1% were large. Changes in statistical significance occurred in 6.5–38.6% of meta-analyses. Changes from significant to nonsignificant were most common (2.1–13.7% meta-analyses). We found no evidence of bias with any rapid review method. Conclusion Searching PubMed only might be considered where a ∼10% risk of the primary outcome OR changing by >20% could be tolerated. This could be the case in scoping reviews, resource limitation, or where syntheses are needed urgently. Other situations, such as clinical guidelines and regulatory decisions, favor more comprehensive systematic review methods.

This article is part of a series featured from the Catalogue of Bias introduced in this volume of Evidence-Based Medicine that describes biases and outlines their potential impact in research studies. Observer bias is systematic discrepancy from the truth during the process of observing and recording information for a study. Many healthcare observations are at risk of this bias. Evidence shows that treatment effect estimates can be exaggerated by a third to two-thirds in the presence of observer bias in outcome assessment. Preventing observer bias involves proper masking in intervention studies including the use of matched placebo interventions where appropriate and training of observers to make assessment consistent and reduce biases resulting from conscious or unconscious prejudices. Where observers are involved in a research study, it is probably not possible for the study to be entirely free of observer biases.

General practitioners generate many clinical questions during consultations. However, when they seek answers to these queries they tend to rely on colleagues and "desk top" references rather than searching the literature themselves. [1][2][3] ATTRACT was created in 1997 to provide rapid, evidence based summaries to clinical queries. All general practitioners in Gwent were invited to send their clinical queries to ATTRACT. For each query an information manager (JB) undertook a rapid search of the literature. The information found was appraised, summarised onto one side of A4 paper, and faxed to the requestor within six hours (see appendix on the BMJ's website for more details on the search, appraisal, and summary process). Examples of the questions received include "Do decongestants help in people with eustachian tube dysfunction?" and "What are the risks of flying while pregnant?" (see BMJ's website for list of most popular questions). We report here an initial evaluation of this service. Participants, methods, and resultsWe developed a brief anonymised questionnaire to seek doctors' views on the usefulness of ATTRACT. We sent this to the first 15 general practitioners to use the service and, one year after the initial survey, to the general practitioners who had asked the 35 most recent questions. In addition, we analysed the clinical queries received during that period (1 January 1997 to 31 January 1998).Forty two (84%) of the 50 general practitioners replied. Of these, 29 (69%) rated the service "very useful" and 13 (31%) rated it "useful." All the respondents rated the service as "very quick" or "quick," and all reported that they would use it again. For the 40 doctors who replied to the question about the effect of the supplied information, nine clinicians were already practising in line with the evidence supplied, 24 changed their practice as a result of the information, six did not change practice because of the weakness of the presented evidence, and the remaining one asked a question about prognosis so that change in practice was not relevant.Of the 193 questions received by ATTRACT in 13 months, 124 (64%) related to therapeutic issues and 23 (12%) related to harm, the next most prevalent category (figure). CommentOur study shows that Gwent general practitioners use and value the ATTRACT service. Although our assessment is based on a small number of users, it merits reporting because of the strength of the opinions expressed and our clear impression that fast answers to clinical questions actually change practice. This study is the first to evaluate a fast, evidence based query answering service for doctors in the NHS, and our results support McColl et al's conclusion that doctors want summaries of evidence rather than the skills to produce them themselves. 4 Our results also confirm Smith's and Ely et al's reports that most clinical queries relate to therapeutic issues. The reported changes in practice are encouraging. This is probably because, instead of "pushing" information towards clinicians in th...