Abstract
BACKGROUND
Western industrialised nations face a large increase in the number of older people. People over the age of 60 years account for almost half of the 16.8 million hospital admissions in England from 2009 to 2010. During 2009-10, respiratory infections accounted for approximately 1 in 30 hospital admissions and 1 in 20 of the 51.5 million bed-days.
OBJECTIVE
To determine the diagnostic accuracy and clinical effectiveness and cost-effectiveness of rapid molecular and near-patient diagnostic tests for influenza, respiratory syncytial virus (RSV) and Streptococcus pneumoniae infections in comparison with traditional laboratory culture.
METHODS
We carried out a randomised controlled trial (RCT) to evaluate impact on prescribing and clinical outcomes of point-of-care tests (POCTs) for influenza A and B and pneumococcal infection, reverse transcriptase-polymerase chain reaction (RT-PCR) tests for influenza A and B and RSV A and B, and conventional culture for these pathogens. We evaluated diagnostic accuracy of POCTs for influenza and pneumococcal infection, RT-PCR for influenza and sputum culture for S. pneumoniae using samples collected during the RCT. We did a systematic review and meta-analysis of POCTs for influenza A and B. We evaluated ease and speed of use of each test, process outcomes and cost-effectiveness.
RESULTS
There was no evidence of association between diagnostic group and prescribing or clinical outcomes. Using PCR as 'gold standard', Quidel Influenza A + B POCT detected 24.4% [95% confidence interval (CI) 16.0% to 34.6%] of influenza infections (specificity 99.7%, 95% CI 99.2% to 99.9%); viral culture detected 21.6% (95% CI 13.5% to 31.6%; specificity 99.8%, 95% CI 99.4% to 100%). Using blood culture as 'gold standard', BinaxNOW pneumococcal POCT detected 57.1% (95% CI 18.4% to 90.1%) of pneumococcal infections (specificity 92.5%; 95% CI 90.6% to 94.1%); sputum culture detected 100% (95% CI 2.5% to 100%; specificity 97.2%, 95% CI 94.3% to 98.9%). Overall, pooled estimates of sensitivity and specificity of POCTs for influenza from the literature were 74% (95% CI 67% to 80%) and 99% (95% CI 98% to 99%), respectively. Median intervals from specimen collection to test result were 15 minutes [interquartile range (IQR) 10-23 minutes) for Quidel Influenza A + B POCT, 20 minutes (IQR 15-30 minutes) for BinaxNOW pneumococcal POCT, 50.8 hours (IQR 44.3-92.6 hours) for semi-nested conventional PCR, 29.2 hours (IQR 26-46.9 hours) for real-time PCR, 629.6 hours (IQR 262.5-846.7 hours) for culture of influenza and 84.4 hours (IQR 70.7-137.8 hours) and 71.4 hours (IQR 69.15-84.0 hours) for culture of S. pneumoniae in blood and sputum, respectively. Both POCTs were rated straightforward and undemanding; blood culture was moderately complex and all other tests were complex. Costs and quality-adjusted life-years (QALYs) of each diagnostic strategy were similar. Incrementally, PCR was most cost-effective (78.3% probability at a willingness to pay of £20,000/QALY). Few patients were admitted within a timescale conducive to treatment with a neuraminidase inhibitor according to National Institute for Health and Care Excellence guidance.
LIMITATIONS
The accuracy study was limited by inadequate gold standards.
CONCLUSIONS
All tests had limitations. We found no evidence that POCTs for influenza or S. pneumoniae, or PCR for influenza or RSV influenced antimicrobial prescribing or clinical outcomes. The total costs and QALYs of each diagnostic strategy were similar, although, incrementally, PCR was the most cost-effective strategy. The analysis does not support routine use of POCTs for either influenza or pneumococcal antigen for adults presenting with acute cardiopulmonary conditions, but suggests that conventional viral culture for clinical diagnosis should be replaced by PCR.
TRIAL REGISTRATION
Current Controlled Trials ISRCTN21521552.
FUNDING
This project was funded by the NIHR Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 18, No. 36. See the NIHR Journals Library website for further project information.
Full text of this article can be found in Bookshelf.