Hay fever index

Spring is coming to an end and summer is quickly approaching. Nature is blooming and there are plenty of grass pollen, too. People who suffer from hay fever sneeze, have blocked noses and itchy eyes as an allergic reaction to the pollen from trees and grasses which are released into the air. As more and more people are suffering from allergies including hay fever, this is becoming a growing problem in society. More insight is needed.

Existing sources such as the Hay Fever Radar and surveillance data from Dutch health care research institute NIVEL do provide insight into pollen flows and number of hay fever patients who visit a GP.
Not everyone who has hay fever related complaints will visit a GP and a pharmacy. They may also go to the local drugstore for self-medication products. This could be medication in the form of tablets, nose sprays or eye drops which curb the allergic reaction.
For this publication, the CBS Centre for Big Data Statistics (CBDS) studied whether weekly sales data on hay fever medication at pharmacies could be used as a new source in order to provide an overview of how the hay fever season progresses.

Increased sales

In the figure below, the turnover index for hay fever medication shows the development of weekly sales of hay fever medication relative to the average weekly turnover in 2016. The same figure displays the changing count of allergenic pollen in the air (source: Pollen counts at Elkerliek Hospital) It also shows the pollen counts of tree and plant varieties which trigger moderate, severe or very severe allergic reactions, such as the hazelnut tree, alder, birch, Poaceae, Rumex, Plantago, Cedar of Lebanon, Artemisia and Ambrosia.

Early in spring around the beginning of April, an increase in sales of hay fever medication can be seen. This is also the period in which pollen of certain trees such as birches trigger allergic reactions. Grass pollen appear later in the season, which is also the time of year when the highest sales peak is recorded. Although the grass pollen count is not extremely high, many people suffering from hay fever are sensitive to grass pollen, which may explain the sales peak. Sales figures are therefore not always directly related to the pollen count in the air; sales are primarily related to the severity of the symptoms experienced by hay fever sufferers.

Figure 2 shows that the hay fever season started slightly later this year compared to 2017. Interestingly, a high sales peak was nevertheless already recorded in April of this year. GP visits on account of hay fever are displaying a similar trend (source: NIVEL surveillance). This peak may be related to the warm weather last April, which caused the release of many tree pollen into the air.

Last-minute purchase

Data further show that people tend to purchase medication only when the pollen are released into the air, rather than stocking up on medication beforehand. It does not look like people only purchase medication at the beginning of the season for long-term use. In general, there is a clear trend that when pollen concentrations increase, sales increase as well, even later on in the season. In short, sales data of hay fever medication obtained from chemists seem to provide a fairly accurate picture of the duration and intensity of the hay fever season.

Explanations

CBS obtains transaction data from a large portion of the pharmacies in the Netherlands. The data are more commonly used as input for the consumer price index. In this collaboration, pharmacies have indicated that their data may also be used for statistical research purposes. Weekly sales data of around 30-40 different products used to alleviate hay fever symptoms have been used to produce the hay fever index. The turnover index is not intended to show the total sales volume of hay fever medication in the Netherlands, but to provide more insight into the development of the hay fever season and times when consumers purchase medication.

Feedback: 

We would like to know your opinion on this beta product and its potential applications. We are open to ideas on how this social phenomenon can be described more accurately using new data sources. Please use the form below to submit your feedback.