June saw a sharp rise in the number of daily confirmed cases in Sweden, with more than 1,000 cases reported on half the days that month, a sharp rise from previous high points of 700. This was put down to an increase in testing as coronavirus tests were made available to anyone with symptoms, regardless of their job or need for hospital care, many weeks after this was the case in countries like Germany, France and Sweden's Nordic neighbours.

While the rise was expected after the change to testing policy, what might seem more surprising is the rapid fall in daily confirmed cases throughout July, with fewer than 500 newly confirmed cases each day in the second half of the month.

This marked drop in the infection rate has seen Sweden at last removed from several countries' travel ban lists, including Germany and Switzerland, while people from a growing number of Swedish regions each week are being allowed in to Denmark and Norway.

One potential explanation for this sharp drop is that the increase in June represented a testing backlog.

“The number of positive tests has decreased, especially if you take into account that the number of tests performed has increased substantially over the past month or so. It's also well known that PCR tests for an ongoing infection can be positive quite long after you had an active infection, even one or two months after,” explains Jan Albert, a professor and senior consultant in infectious disease control and virology.”

He also notes that the general downward trend is clear also when looking at coronavirus-linked intensive care admissions and deaths, two of the most trustworthy data, as well as excess mortality across the country as a whole, which can be measured as a way of tracking both coronavirus-related deaths that are not recorded as such, and the impact of the pandemic on public health overall.

Meanwhile, the data for confirmed cases can be harder to interpret as testing policy has varied over time and within regions, with Sweden struggling for months to provide sufficient capacity.

“We had a substantial spread in society, resulting in severe cases and deaths, primarily in March and April. There is a lag between becoming infected and being admitted to an ICU and possibly dying, it's expected that these numbers take time to reflect the changes,” explains Albert.

The graphs above are based on Public Health Agency data from July 24th, and the figures for deaths are based on the date the death was recorded.

Albert sees three likely factors behind the sharp fall in the infection rate, although he stresses the difficulty of knowing how these factors interacted and what weight each one had.

One factor is public adherence to guidelines such as social distancing and good hand hygiene, known to reduce the spread of infectious diseases. Although surveys by Sweden's Civil Contingency Agency suggest that people are over recent months following social distancing to a lesser extent, and authorities have spoken in the bi-weekly press conferences of concerns over crowding at popular summer destinations, it seems that people have not relaxed to the extent that it has reversed the downward trend.

“Distancing hasn't increased dramatically over the last few weeks or months – rather the opposite I would say. But the impact of social distancing and other measures of a partial lockdown come gradually, things that started in March and April continue into May and June so we may be seeing the effect of that now. Social distancing is not perfect but it exists; some people don't adhere to the recommendations but many do, especially the more vulnerable groups,” says Albert.

A busy scene by the water in Malmö in late June. Photo: Johan Nilsson / TT

The summer season also has an impact on behaviour that may be beneficial to disease control, and Albert says this effect may be especially pronounced in Sweden where it's common to take several consecutive weeks' vacation.

“Schools are closed, daycares have fewer children attending, people are on vacation, people spend more time outdoors or in rural areas and with fewer people,” he notes.

The summer may also have an effect on the way the virus behaves. Albert says: “Viruses of this type – although it's not been proved to be the case for Covid-19 – tend to spread more effectively in a country like Sweden in winter, due to physiological things, the droplets that are generated when people are sick are smaller and fly better in cold and dry climates.”

And a final possible factor for the sharp fall in infections is what the professor says is the “most difficult factor”: the question of immunity.

Sweden's health authorities have denied that their response focuses on herd immunity as an aim. There is currently no scientific consensus on what level of immunity exposure to Covid-19 provides, or how long it may last, and it's hard to measure immunity.

But Albert says the figures do appear to support the Public Health Agency's theory that a significant proportion of people have some immunity, thus slowing the spread of the virus.

“My guess, although it's only a guess, is that we do see an effect of immunity of some type even if it's not revealed in the proportion of people testing positive for antibodies,” he says.

What can we expect next?

The impossibility of knowing which of these factors contributed to the fall in infections makes it hard to make any predictions for autumn or to make changes to the health guidelines in place. Without knowing how crucial measures such as social distancing are, and whether immunity has played a role and whether this will continue to be the case, it is tough to adjust the strategy safely.

Albert says it's hard to know what will happen in autumn, but that an uptick in transmissions – a so-called second peak – is likely.

“It's really difficult to make decisions on things like the choice between a full lockdown, partial lockdown or no lockdown. No one really knows what's the best long-term strategy and that's still true, we don't know what would happen if we stopped the partial lockdown and social distancing – whether there'd be a dramatic increase in the number of new transmissions or if it would be manageable,” he says.

“I think the safest route is to continue with the recommendations and pay close attention to any increases, try to detect clusters of transmission early and break transmission chains, which is what's happening.”

“We don't know everything, we can't say we're past the worst of this so we need to keep up testing, things like social distancing, and contact tracing, so that we don't have a dramatic increase.”

A woman wears a facemask at a bus stop in Stockholm. Photo: Stina Stjernkvist/TT

What's more, the situation is still subject to regional variations.

The figures are moving in the right direction on a national level and in the regions which were hard hit early on, including Stockholm and Sörmland, but some regions are at a plateau stage rather than a downturn. 

The Public Health Agency has warned that localised outbreaks remain likely, as has been the case in other countries even with a far lower spread of infection than Sweden, such as Germany and Spain. One such outbreak occurred in Gällivare in northern Sweden.

• Does a town in northern Sweden really have an 'uncontrollable spread' of Covid-19 infections?

“We have quite a quick positive development of the pandemic in Sweden – which shouldn't be taken as a reason that we can live as normal again. This depends on us continuing to do social distancing and other measures in a good way,” state epidemiologist Anders Tegnell said at Thursday's press conference.

And despite the positive signs over recent weeks, the Swedish response will still be under scrutiny for a long time to come, with a government commission to look into the strategy and assess whether deaths could have been avoided.

Since the start of the outbreak, over 5,600 people have died after testing positive for the virus, and many hundreds of thousands more have likely been infected by the coronavirus. The majority of patients who catch the virus suffer only mild symptoms and recover, but there are many reports of long-term effects, something else Swedish authorities are looking into.