We were notified last week of a covid-19 case in my seven-year-old daughter’s class. My wife and I both work in the NHS, but we still had to recheck the latest school and government guidance. In the absence of persistent cough, fever, or loss of taste/smell, children can continue to attend school. Other than the use of alcohol hand gel at the school gates, external signs of the pandemic are few, which is positive for pupils who have to carry on learning and developing. Meanwhile, increasing cases, admissions, and deaths add to evidence of acute and longer-term effects of SARS-CoV-2 infection in children. “Cognitive dissonance”, the theory describing the “discomfort people feel when two cognitions, or a cognition and a behaviour, contradict each other,” has been invoked to explain contradictory individual-level behaviours during the pandemic. Perhaps surprisingly, it is seldom used to describe situations arising in healthcare research and policy, where it may be pertinent.

Pre-print culture has mushroomed since early 2020, facilitating early, open scientific debate. Some journals, including those with highest impact factors, do not always respond favourably when manuscripts have been posted as pre-print and already covered in media—whether social or mainstream—despite having processes of submission, review, and publication lasting several months. A recent preprint on medRxiv which suggested a considerable risk of myocarditis with covid-19 vaccination was found to have made a calculation error, and retracted. The un-reviewed article may have contributed to the delayed decision to vaccinate 12-15 year olds in the UK and has been used by those opposed to vaccination. “Living” systematic reviews aim to collate evidence relating to particular research questions in an updatable manner, but the scale and speed of research makes this challenging. For example, a living systematic review of school transmission of SARS-CoV-2 was published in December 2020, finding poor quality evidence to support transmission in children in schools, but has not been updated online since. Part of the solution is to understand that, although the pace of covid-related research is unprecedented, publication and dissemination models are not always quick enough for public health need, and to work towards scientific models which put patients rather than publications first.  

Long covid is a new disease being defined at the same time as it is treated and managed. The individual, population, and economic burden has been underplayed due to a focus on acute hospitalisation and mortality. Traditional evidence takes time to accumulate, highlighting the importance of patients’ experience and the expertise of frontline clinicians in policy and system responses. However, this is not necessarily reflected in the membership of advisory committees and decision-making organisations, or research funding calls which have prioritised mechanism and pathophysiology above design and evaluation of clinical services. This creates potential obstacles to recognition and awareness of long covid, and delayed selection and evaluation of treatment strategies. The way forward must involve acceptance of limitations of current review panels and governance structures to deal with new diseases of this scale, and enabling greater cross-pollination between research, clinical, public health, and policy spheres. A change in the way patient and public involvement is viewed and enacted is also long overdue.

In the age of social media, open science, and the continuous news cycles, we are all looking at other countries and other settings as much as our own. Given news and research from China and Italy in early 2020, one could argue that the first UK lockdown was delayed, but the second lockdown was definitely late judged by the projected impact of covid-19 on high-risk populations, subsequent excess mortality, and indirect effects on non-covid care. Approaches to masks, opening of schools and universities, and border controls are other instances where there has been a difference between published evidence or policy in other countries, and the UK situation. The eventual decision to vaccinate children in the UK occurred a few weeks after school started, when cases had already started to rise. Several other countries had already begun vaccination programmes in those over the age of 12 and 16 years with published data. Delays, differences, or discrepancies in health policy and the rationale for decision-making should be clearly communicated to the public.

The common thread is the gap in clear, public communication of information, and acknowledgment of the limitations and public involvement in decision-making. Cognitive dissonance may be likely or even inevitable in pandemics, but reduction strategies can include acquiring new information; changing beliefs, attitudes, or behaviours; or reducing the importance of the dissonant changing beliefs, attitudes, or behaviours. This framework may help public, researchers, health professionals and policymakers navigate tensions in the rapidly-shifting evidence and policy landscape of emergency preparedness.

Amitava Banerjee, Professor of Clinical Data Science and Honorary Consultant Cardiologist, Institute of Health Informatics, University College London.

Twitter: @amibanerjee1

Competing interests: none declared.