Dec 15, 2011 (CIDRAP News) – The use of statins—widely used lipid-control drugs—was associated with a 41% lower death rate in patients who were hospitalized with influenza, according to a surveillance study from the 2007-08 flu season that spanned 10 states.
Over the past several years, researchers have identified possible statin benefits for other diseases. For flu, the drugs have the potential to inhibit the release of cytokines, pro-inflammatory chemicals that are thought to play a role in the type of severe pathophysiologic changes in seen in human H5N1 avian flu infections.
Studies to gauge the effect of statins on flu have produced mixed results. But the latest study, published yesterday in the Journal of Infectious Diseases (JID), is the first to limit the analysis to lab-confirmed flu cases.
The analysis is based on data from the influenza hospitalization surveillance system of the US Center for Disease Control and Prevention's Emerging Infections Program, which encompasses 59 counties in 10 states. Demographic and clinical data were collected from each patient's medical record, and the researchers obtained flu vaccination information from charts, registries, physicians, or patients.
The researchers recorded any statin use before or during hospitalization, but they did not collect the dose or frequency.
Surveillance subjects included adults who were hospitalized between Oct 1, 2007, and Apr 30, 2008, within 14 days of a positive influenza test. Testing methods included viral culture, immunofluorescence antibody staining, reverse-transcriptase polymerase chain reaction (PCR), a commercial rapid flu test, serologic testing, or an unspecified test noted in the patient's records.
With 3,043 patients included in the analysis, the average age was 70.4 years, and 56% were women. More than half (57.1%) had been vaccinated, and 33.3% (1,013) were given statins before or during hospitalization. Of that group, 76% were on statins before and during hospitalization. Those who used statins were more likely to be male and white, to have a cardiac or other underlying medical condition, and to have been vaccinated against flu.
Of the study group, 151 (5%) died within 30 days after their flu diagnosis, the majority of them shortly after hospital discharge.
After controlling for age, race, underlying medical conditions, vaccination status, and antiviral use, the researchers reported a 41% reduction (adjusted odds ratio, 0.59; 95% confidence interval, 0.38-0.92) in mortality in patients who took statins before or during hospitalization.
The results agree with the findings of two other studies that explored links between statins and flu mortality, but those studies used broader disease classifications that included pneumonia, rather than lab-confirmed flu, according to the authors.
However, they said a 2010 study found no benefits from statin use for acute respiratory infections in a primary care setting. That study didn't assess prescription adherence and used different disease end points that didn't include more severe outcomes. "Statins may be more beneficial at preventing disease outcomes, such as death, but may not play a role in reducing the incidence of infection or minor illness," the authors wrote.
Several limitations cited
The researchers detailed three main limitations of their findings: testing issues, limitations of chart data, and the possibility of a "healthy user bias" in people who take statins.
They noted that the study cohort may not reflect all patients hospitalized with flu, including those who weren't tested for the disease. Also, the limited sensitivity of the rapid tests could lead to underestimation of the number of cases.
According to the CDC, the sensitivity of rapid flu tests ranges from 50% to 70%, and their accuracy is influenced by the level of flu in the community. False-negatives are more likely when disease prevalence is high.
Ann Thomas, MD, MPH, the study's corresponding author and a public health physician with the Oregon Public Health Division, in an e-mail detailed the percentages for the flu confirmation methods: 75% rapid test, 11% direct immunofluorescent assay (DFA), 11% cell culture, and 6% PCR. She said the total exceeds 100% because some patients with positive rapid tests likely had other confirmatory tests.
She said that, since the 2009 H1N1 pandemic, clinicians have become more interested in knowing the subtype, and PCR has now become the dominant test to identify flu cases for the Emerging Infections Program surveillance system. However, before that time, including the 2007-08 flu season, about 75% of the surveillance systems cases were diagnosed by rapid test.
She noted that most of the cases during the study were identified during peak flu season, when the positive predictive value of rapid testing is higher, which Thomas said makes the study group hopeful that the test results are accurate.
It's unlikely that statin use would be associated with false-positive rapid test results or that death is more or less likely in subjects with false-positive results, she wrote. "My guess would be that inclusion of patients who did not really have influenza would bias us towards the null," she said
Exploring statin use in patients who had negative rapid tests would require the investigators to do chart abstractions, which they weren't able to do because they didn't collect information on that group, Thomas noted.
Another challenge the researchers had was determining the true number of deaths. Since the study relied mainly on chart data, they attempted to clarify the number of deaths by reviewing local vital statistics records to see if any of the hospitalized patients died within 30 days of hospital discharge.
The reliance on chart data also made it difficult for the researchers to assess the underlying functional health status of the patients, which they wrote could influence the relationship between statins and flu severity, introducing a possible "healthy user" bias. However, they noted that the statin users were older, had more underlying conditions, and had longer hospital stays, suggesting that their baseline health was no better than that of patients not taking the drugs.
The researchers concluded that their findings suggest that statins are a promising area for further study and could be a useful adjunct to antiviral medication and vaccines, especially when viruses aren't susceptible to medication or when the flu vaccine is in short supply or not well matched to the circulating strains.
Future studies—ideally randomized controlled trials—exploring a possible role for statins in treating influenza should include analysis of functional status, dose and duration of therapy, statin use in younger patients, and identification of the most effective class of statins, they wrote.
Slowly mounting evidence
In an editorial in JID yesterday, Dr Edward Walsh, with the Infectious Diseases Division at the University of Rochester School of Medicine in New York, wrote that the findings are a significant addition to the slowly mounting evidence that statins may reduce annual flu burden and deaths, because the study links deaths to lab-confirmed flu.
As in many observational studies, unrecognized factors, such as the "healthy user" bias, may affect the results and conclusion, Walsh wrote, noting that similar concerns have been raised about flu vaccine efficacy in older people.
The analysis did not find that the influenza vaccine and antivirals reduced the number of deaths, though it did show a nonsignificant benefit for antivirals, he pointed out.
Walsh said the findings raise several key questions about the relationship between statins and flu deaths, such as what mechanism is responsible. For example, he said it's unclear if statins are more effective in flu complicated by bacterial infection or if the main benefit is due to anti-inflammatory effects that help prevent heart attacks and strokes during flu epidemics. It's often difficult to assess the cause of deaths in chart review studies, so scientists can only speculate about what statin mechanisms might provide a benefit, he wrote.
More high-quality prospective observational studies are needed to confirm the findings, but a randomized trial of the effects of long-term statin use on flu outcomes would be logistically difficult, and clinicians probably wouldn't administer statins solely to reduce flu severity, Walsh said.
However, he said such studies could support the use of statins in a pandemic setting, a strategy suggested in 2006 by David Fedson, MD, an internationally known vaccine expert. Fedson has said that if research confirms the benefits of statins for flu patients, they may offer hope for patients in countries where antiviral and vaccine supplies are scarce.
Vandermeer ML, Thomas AR, Kamimoto L, et al. Association between use of statins and mortality among patients hospitalized with laboratory-confirmed influenza virus infections: a multistate study. J Infect Dis 2011 Dec 14 [Abstract]
Walsh E. Statins and influenza: can we move forward? (Editorial) J Infect Dis 2011 Dec 14 [Extract]
Jun 30, 2006, CIDRAP News story "Paris meeting airs avian flu impacts, possible treatment"