Evolving concerns about the COVID-19 pandemic: A content analysis of free-text reports from the UK COVID-19 Public Experiences (COPE) study cohort over a two-year period

This longitudinal content analysis of over 41,000 free-text reports from the UK COPE study reveals that public concerns about COVID-19 evolved significantly over two years, shifting from a primary focus on harming others to a greater emphasis on personal harm, while overall concern levels declined and reflected changing perceptions of risk and responsibility.

The Gist

Imagine the COVID-19 pandemic as a massive, two-year-long storm that swept across the UK. This paper is like a detailed weather report, but instead of measuring wind speed and rain, the researchers measured what was on people's minds.

They asked over 11,000 people, "What are you most worried about?" at five different points in time, from the very first lockdown in 2020 to the point where restrictions were lifted in 2022. They analyzed over 41,000 handwritten answers to see how the "weather" of public fear changed.

Here is the story of those two years, told through simple analogies:

1. The Big Shift: From "Saving Others" to "Saving Myself"

Think of the pandemic as a relay race where the baton of worry was passed around.

• The Early Days (2020): At the start, the baton was heavy with altruism. People were mostly worried about others. They were terrified of catching the virus and passing it to their grandparents, their neighbors, or the vulnerable people in their community. It was like everyone was holding an umbrella, trying to keep their whole village dry.
• The Later Days (2022): As the race went on, the baton shifted. By the end, people were mostly worried about themselves. The focus turned to "Will I get Long COVID?" or "Will I get sick again?" The umbrella was now held tight over just one person's head.

The Takeaway: In the beginning, people were driven by love for their community. By the end, they were driven by self-preservation.

2. The Six Buckets of Worry

The researchers sorted all the fears into six big buckets. Here is what was in them and how they changed:

• Bucket 1: Personal Harm (The "Me" Bucket)

  • What it was: Fear of getting sick, dying, or getting "Long COVID" (a lingering illness).
  • The Change: This bucket got heavier over time. Even though vaccines arrived and the virus seemed less deadly to the average person, people who had health conditions or were older felt more exposed as rules were relaxed. They felt like they were walking a tightrope without a safety net.

• Bucket 2: Harm to Others (The "We" Bucket)

  • What it was: Fear of infecting family or the vulnerable.
  • The Change: This bucket got lighter. As people got vaccinated and the virus became more common, the panic of "I will kill my grandma if I sneeze" faded, replaced by the thought, "Well, we're all getting it anyway."

• Bucket 3: Social & Economic Impact (The "Life" Bucket)

  • What it was: Fear of losing jobs, schools closing, or the economy crashing.
  • The Change: This bucket was full at the start (panic buying, lockdowns) but gradually emptied out. As people got used to the new normal, the fear of the economy collapsing became less of a daily nightmare, though the scars remained.

• Bucket 4: Stopping the Spread (The "Rules" Bucket)

  • What it was: Worrying about whether masks, tests, and vaccines would work, or if people were following the rules.
  • The Change: This became a major source of frustration. People started feeling like they were the only ones holding the door shut while everyone else was kicking it open. The fear wasn't just the virus; it was the other people not following the rules.

• Bucket 5: Government & Media (The "Trust" Bucket)

  • What it was: Confusion about what the news said, distrust in politicians, and feeling like the rules changed too often.
  • The Change: This was a constant background noise. People felt like they were trying to drive a car with a foggy windshield and a GPS that kept changing the destination. High-profile politicians breaking the rules made people feel like the "rules" were a joke, which made them less willing to follow them.

• Bucket 6: General Uncertainty (The "Unknown" Bucket)

  • What it was: "Will this ever end?" or "What is going to happen next?"
  • The Change: This faded away. The fog lifted. People knew the virus wasn't going away, but they stopped worrying about the unknown and started worrying about the known (like getting sick again).

3. The "Pandemic Fatigue" Effect

Imagine you are holding a heavy box for two years. At first, you are terrified you'll drop it. But after a while, you just get tired. You put it down, even if it's still heavy.

The study found that by the 24-month mark, fewer people were reporting worries overall. This wasn't necessarily because the virus was gone; it was because people were exhausted. They had to make a choice: keep worrying and stay in lockdown forever, or accept the risk and try to live their lives. Most chose the latter.

4. Why This Matters for the Future

The authors say that if we have another pandemic, we can't just ask, "Are you scared?" We need to ask, "What kind of scared are you?"

• Early on: We need to tell people, "Protect your neighbors."
• Later on: We need to tell people, "Here is how to protect yourself, and here is why the rules are fair."

In a nutshell:
This paper shows that human fear is like a chameleon. It changes color depending on the situation. At the start of the pandemic, we were a community of protectors. By the end, we were a group of individuals trying to navigate a world that felt less safe but also less restrictive. To handle future crises, leaders need to understand that the "fear" of the public isn't a single thing—it's a complex mix of love for others, fear for ourselves, and frustration with the rules.

Technical Summary

1. Problem Statement

During the COVID-19 pandemic, risk perception was a primary driver of infection-prevention behaviors. While quantitative studies have tracked general trends in fear and susceptibility, they often fail to capture the specific, nuanced nature of public concerns or how these concerns evolve over time. Existing research typically relies on broad metrics (e.g., "fear of infection") rather than granular qualitative data. There is a critical gap in understanding how specific concerns—ranging from personal health to economic impact and trust in institutions—shifted in frequency and priority across the population during the first two years of the pandemic. This study aims to fill that gap by analyzing open-ended qualitative responses to understand the dynamic nature of public risk appraisal.

2. Methodology

• Study Design: A longitudinal content analysis of free-text responses from the UK COVID-19 Public Experiences (COPE) study, a prospective mixed-methods cohort study.
• Data Source:
  • Cohort: 11,113 adults recruited at baseline (March/April 2020) via the HealthWise Wales registry and social media.
  • Time Points: Five survey waves conducted over 24 months: Baseline (Lockdown 1), 3 months, 12 months, 18 months, and 24 months.
  • Prompt: Participants were asked: "If you are worried about coronavirus (COVID-19), please tell us what your main concerns are."
  • Volume: 41,568 open-text responses were analyzed.
• Analytical Approach:
  • Theoretical Framework: The analysis was guided by the COM-B model (Capability, Opportunity, Motivation – Behavior) and PRIME Theory (Plans, Responses, Impulses, Motives, Evaluations), focusing on the "affective" (emotional/automatic) component of motivation.
  • Coding Process: An inductive-deductive hybrid approach was used. An initial coding framework was developed from baseline data and applied deductively to follow-up waves, with new categories added inductively as the pandemic evolved.
  • Team: 11 researchers coded the data in Microsoft Excel. Inter-rater reliability was checked on 10% of baseline data.
  • Categorization: Responses were coded into six main categories and 30 sub-categories.

3. Key Contributions

• Granular Evolution of Concerns: Unlike studies relying on Likert scales, this paper provides a high-resolution map of how specific concerns (e.g., "Long-Covid," "NHS capacity," "Government inconsistency") emerged, peaked, and declined over 24 months.
• Shift from Pro-Social to Self-Focused: The study identifies a distinct temporal shift in the type of concern, moving from collective responsibility (harm to others) in the early stages to individual vulnerability and frustration with non-compliance in later stages.
• Integration of Theory and Qualitative Data: It successfully applies PRIME Theory to explain how competing impulses (e.g., fear of infection vs. fear of economic ruin) balanced to influence behavior at different pandemic stages.
• Large-Scale Qualitative Analysis: The analysis of over 41,000 free-text comments represents one of the largest qualitative content analyses of pandemic risk perception to date.

4. Key Results

Overall Trends:

• The proportion of respondents reporting any concern declined from 75.3% at baseline to 65.8% at 24 months.
• Personal Harm concerns increased in relative frequency (41.8% at baseline $\rightarrow$ 52.7% at 24 months).
• Harm to Others concerns decreased significantly (48.5% at baseline $\rightarrow$ 28.6% at 24 months).

Category-Specific Findings:

1. Personal Harm:
   • Initially focused on severe illness, hospitalization, and death.
   • Over time, concerns shifted toward clinical vulnerability (peaking at 3 and 24 months) and Long-Covid (rising from 0.4% to 13.7%).
   • Concerns about the inability to care for dependents (children, elderly) remained persistent.
2. Harm to Others:
   • Dominated early responses, focusing on family members and the clinically vulnerable.
   • Declined as vaccination rates rose and restrictions eased, though concerns about transmitting the virus to vulnerable relatives persisted.
3. Social and Economic Impacts:
   • High at baseline (26.7%), focusing on panic buying, NHS capacity, and financial insecurity.
   • Declined steadily over time (10.4% at 24 months), though mental health and isolation remained sub-themes.
4. Preventing Spread:
   • Increased over time (9.8% $\rightarrow$ 19.3%).
   • Shifted from lack of testing/vaccines (early) to non-compliance by others, lack of workplace protection, and anxiety about new variants (Omicron) and the removal of restrictions.
5. Government and Media:
   • Consistently low in frequency (~3-4%) but high in intensity.
   • Key themes: Trust erosion due to perceived inconsistencies (e.g., Dominic Cummings scandal), lack of cohesion between UK nations, and "scaremongering" or contradictory media coverage.
6. General Concerns:
   • Focused on uncertainty about the future and the duration of the pandemic; these decreased as the situation became more "endemic."

5. Significance and Implications

• Public Health Communication: The study highlights that a "one-size-fits-all" message is ineffective. Communication strategies must adapt to the evolving dominant concern (e.g., shifting from "protect others" to "protect yourself from Long-Covid" as the pandemic progressed).
• Behavioral Intervention: Understanding the balance between competing concerns (e.g., fear of infection vs. fear of economic loss) is crucial for designing interventions that do not trigger fatalism or rebellion.
• Trust and Compliance: The findings underscore that trust in government and media is a critical determinant of adherence. Perceived injustice (e.g., elites breaking rules) and inconsistent messaging significantly undermine public motivation to follow guidelines.
• Future Pandemic Planning: The authors argue for the necessity of longitudinal, mixed-method monitoring of public concerns to detect emerging issues (like Long-Covid or vaccine hesitancy) in real-time, allowing for agile policy adjustments.

Limitations: The cohort was self-selecting, with an over-representation of older adults, females, and those with pre-existing conditions, and an under-representation of ethnic minorities. Consequently, the findings reflect the experiences of this specific demographic rather than the entire UK population.