Not the most wonderful time of the year: effects of the influenza season on working-age adults with diabetes
While adopted in many countries, there is considerable debate about the usefulness of vaccination against influenza. This debate hinges on two elements. Firstly, the absolute risks of influenza and its sequelae in the groups to be vaccinated; and secondly, the intrinsic efficacy of the vaccine, i.e. the reduction in risk that can be achieved by vaccination. While the intrinsic efficacy of the vaccine is heavily contested there is less controversy about the increased risk of certain patient groups. A recent study in Diabetologia by Darren Lau and colleagues adds further weight to the notion that diabetes patients are at an increased risk for the adverse consequences of influenza.
Since vaccination is already recommended for all elderly (age > 65 years) adults, Lau and colleagues focused on working age (age ≥18 and <65 years) adults with diabetes as a high-risk group relative to those without diabetes.
The study used data from Manitoba, Canada, from 2000 to 2008. All working age adults with diabetes were identified and matched with up to two non-diabetic controls. The rates of physician visits and hospitalisations for influenza-like illness, hospitalisations for pneumonia and influenza, and all-cause hospitalisations were analysed. The study included 163,202 people, mean age 52.5 years, of whom just under half (48.5%) were women. The data showed that adults with diabetes had more co-morbidities and received influenza vaccination more often than those without diabetes. After adjusting for these differences, adults with diabetes still had a 6% greater increase in all-cause hospitalisations associated with influenza compared to adults without diabetes. No statistically significant differences were detected in influenza-attributable rates of the other outcomes, i.e. influenza-like illness or pneumonia and influenza. This suggests that diabetes patients and controls are equally likely to have mild influenza, but that diabetes patients are more prone to progress to the serious manifestations of influenza. In absolute terms, the influenza season was responsible for an additional 6 hospitalisations per 1000 diabetes patients per year, while it did not affect hospitalisation rates in those without diabetes.
In their discussion the authors do some calculations to suggest that vaccination in this group of working-age diabetes patients would be cost-effective if the vaccine prevented 20% or more of hospitalisations. However, this study can not truly answer the question whether vaccination will be useful. As the data from this study show, for every 1000 diabetes patients who are infected with influenza and who get influenza-like illness, only a very small minority of 6 patients will end up in hospital because of it. It is unclear to what extent this specific subset of people would benefit from influenza vaccination. Even if overall efficacy of the vaccine in preventing influenza would be 80%, would these 6 patients be in the group that responded to the vaccine, or in the group that did not respond? It is also unclear how many adverse events would occur if all 1000 patients were vaccinated. Some papers in the British Medical Journal have highlighted all the flaws in the evidence-base for the efficacy of influenza vaccination. While studies such as this new study by Lau are highly useful in identifying the specific harms of influenza, we remain in desperate need of better evidence for vaccination strategies. As the same authors have previously shown in a study in Thorax , the efficacy of vaccination can not be determined from observational data: it is appropriately randomized controlled trials that we need.
^ Lau D, Eurich DT, Majumdar SR, Katz A, Johnson JA. Working-age adults with diabetes experience greater susceptibility to seasonal a population-based cohort study. Diabetologia 2014. DOI: 10.1007/s00125-013-3158-8 available at Diabetologia