Rapid flu tests more accurate when positive, but negative results require confirmation

Rapid influenza diagnostic tests (RIDTs) were more accurate at diagnosing the flu than ruling it out in a new study, indicating that patients with a negative result would need additional testing.

Researchers reviewed 159 published studies to determine the accuracy of RIDTs in diagnosing flu in adults and children presenting with flu-like symptoms. RIDTs were defined as any commercially available assay that identified flu antigens or neuraminidase activity in respiratory specimens through simple immunochromatographic formats. In all studies, RIDTs were tested against one of two accepted reference standards. Results appeared online Feb. 27 at Annals of Internal Medicine.

Specificity was more consistent across studies than sensitivity, with specificity estimates ranging from 50.5% to 100% and sensitivity estimates ranging from 4.4% to 100%. For all RIDTs, the pooled sensitivity from bivariate random-effects regression was 62.3% (95% CI, 57.9% to 66.6%) and the pooled specificity was 98.2% (95% CI, 97.5% to 98.7%). This corresponds to a positive likelihood ratio of 34.5 (95% CI, 23.8 to 45.2) and a negative likelihood ratio of 0.38 (95% CI, 0.34 to 0.43).

Subgroup analyses showed a higher pooled sensitivity in children (66.6%; 95% CI, 61.6% to 71.7%) compared with adults (53.9%; 95% CI, 47.9% to 59.8%; P<0.001). Pooled sensitivity differences between children and adults remained statistically significant when adjusted for brand of RIDT, specimen type or reference standard. Virus type also had an effect on the accuracy of RIDTs. The tests had better sensitivity for detecting influenza A (64.6%; 95% CI, 59.0% to 70.1%) than influenza B (52.2%; 95% CI, 45.0% to 59.3%; P=0.05).

The authors noted that a positive test is unlikely to be a false-positive result, allowing clinicians to confidently diagnosis and treat the flu.

“RIDTs fill a void at the point of care that no other test is likely to fill in the near future: as a first-line test to be confirmed (especially if negative) by more time-consuming, definitive testing,” they wrote. “As long as clinicians understand the limitations of RIDTs, namely that a negative result is unreliable and should be confirmed by using culture or RT-PCR [reverse transcription polymerase chain reaction], RIDTs could enable clinicians to institute prompt infection-control measures, begin antiviral treatment in high-risk populations, and make informed decisions about further diagnostic investigations.”

A second meta-analysis also published by Annals on Feb. 27 concluded that oral oseltamivir and inhaled zanamivir may provide a net benefit over no treatment for influenza in high-risk populations, although quality of evidence was low due to study designs.

Oral oseltamivir compared with no treatment reduced mortality in high-risk populations (odds ratio [OR], 0.23; 95% CI, 0.13 to 0.43), hospitalization (OR, 0.75; 95% CI, 0.66 to 0.89) and duration of symptoms (33 hours; 95% CI, 21 to 45 hours). For every 1,000 patients, approximately 12 patients require hospitalization, and oral oseltamivir treatment can reduce this by 3 to 9 per 1,000 patients. Earlier treatment with oseltamivir was generally associated with better outcomes.

Inhaled zanamivir led to shorter symptom duration (23 hours; 95% CI, 17 to 28 hours) and fewer hospitalizations (OR, 0.66; 95% CI, 0.37 to 1.18) but more complications than no treatment. Direct comparison of oral oseltamivir and inhaled zanamivir does not suggest important differences in key outcomes, the authors said.

Although there is low to very low confidence in the estimates of effect, randomized controlled trials and “the substantial burden of influenza disease worldwide” lends importance to the findings, the authors wrote.

“The potential positive effect of earlier rather than later administration of oseltamivir on death in hospitalized patients and suggestions that pregnant women, children and patients who are immunocompromised may also benefit from treatment are among the key contributions of our study,” they wrote. “In addition, we found moderate quality evidence for the reduction in signs and symptoms from treatment with inhaled zanamivir compared to no treatment.”