Study setting and design
Ulvik is a rural municipality in the innermost part of the Hardanger fjord in Western Norway. It had 1061 inhabitants at the time of the outbreak [7], a single nursing home, one kindergarten, and a primary and secondary school.
The outbreak commenced on January 29th 2021 when the first case was detected. The sampling frequency of nasopharyngeal and throat swabs combined with contact-tracing was ramped up and was freely available to all inhabitants regardless of symptoms. All symptomatic individuals and those who had been in close contact with SARS-CoV-2 positive persons were specifically encouraged to be tested. A total of 809 tests were performed on 554 individuals between January 29th and the outbreak’s culmination on February 22nd. Samples were analyzed by nucleic acid amplification testing (NAAT) at Haukeland University Hospital. The B.1.1.7 variant was detected in all positive samples.
This was an observational study where all 134 persons with a positive NAAT were invited to answer web-based questionnaires two months after testing positive, which was the time that it took to obtain ethical approval and develop the research protocol. Those who responded after two months were invited to the seven-months follow-up.
Definition of study time periods and corresponding data sources
The data from the two questionnaires were conceptually divided into three time periods: The acute phase of infection (I.) was assessed retrospectively in the first questionnaire. Status at two months (II.) and status at seven months after infection (III.) were prospectively assessed in their corresponding questionnaires distributed at two and seven months, respectively.
Main outcome variables
The main outcomes were symptoms during the acute phase, presence of symptoms at two- and seven-months follow-up, and chronic fatigue or insomnia at seven-months follow-up.
I. The acute phase of infection
Participants were asked whether they had COVID-19 symptoms (yes/no/uncertain) during the outbreak. Symptomatic or uncertain persons were asked for the presence of specific symptoms (cough, dyspnea, fever >38 Celsius, muscle pain, joint pain, headache, lethargy, confusion, dizziness, loss of memory, concentration difficulties, loss of taste and smell, loose stools, abdominal pain, chest pain, rash, skin bleeding, tinnitus, loss of hearing and an open-ended question concerning other symptoms). All symptom-related questions had a checkbox as to whether each symptom was still present at the time of answering the questionnaire.
II. Two months follow-up
Participants with symptomatic infection during the acute phase were asked whether they still felt ill at two months when completing the first questionnaire. Answers were dichotomized into symptomatic yes/no, and uncertain participants were categorized as symptomatic at two months if they reported at least one specific symptom and had checked the "still present" checkbox for symptoms during the acute phase, otherwise they were categorized as asymptomatic.
III. Seven months follow-up
All participants, regardless of symptom status during the acute phase and two-months follow-up, were asked whether they experienced any symptoms in relation to the infection (yes, no or uncertain), seven months after. Participants answering yes or uncertain were asked for specific symptoms, omitting the open-ended question. Answers were dichotomized into symptomatic yes/no, with uncertain participants being categorized as currently having symptoms if they reported at least one specific symptom.
Fatigue
All participants were asked a single-item question at the two-months follow-up whether they felt fatigued prior to infection and at both two- and seven-months follow-up whether they currently felt fatigued. If answering yes, participants were asked to quantify the degree of fatigue they experienced during the last 24 hours on a sliding visual analogue scale (VAS) from 0 to 100.
The Chalder fatigue questionnaire (FQ) was included at both follow-ups for participants aged 12 years or older. It consists of 13 items of which 11 examine different physical and mental aspects of fatigue [8]. Answers are given on a 4-item Likert scale: Less than usual (0), no more than usual (1), more than usual (2), much more than usual (3). The scores are summarized to give a ‘fatigue score’ (range 0-33), and dichotomized (0, 0, 1, 1) with ‘fatigue’ defined as a dichotomized score of 4 or more. Severe fatigue is present if in addition the ‘fatigue score’ is 23 or more [9]. The last to items on the FQ address the extent and duration of fatigue, and if the fatigue has lasted at least 6 months, it is defined as chronic fatigue (CF). Missing answers to any of the 11 items were imputed from the mean value of answers among the other respondents for the missing item, if four or less items were missing in total for the respondent in question.
Insomnia
The Bergen Insomnia Scale (BIS) [10] was used to assess insomnia at two- and seven-months follow up for all participants, but included in analyses only for respondents aged 12 years or older. BIS is a 6-item questionnaire addressing sleep (questions 1-4) and tiredness (questions 5-6). Answers are given on a 0-to-7-point scale corresponding to number of days in a week during the previous month for which a given item is applicable for the subject. A score of 3 on at least one of questions 1-4 and questions 5-6 was defined as insomnia. Participants who reported subjective sleep problems and met the BIS criteria were defined as insomnia cases.
Background variables
In the first questionnaire (at two months) participants were asked for their gender, weight and height prior to infection, country of origin and pre-existing comorbidities (diabetes, asthma, allergy, anxiety, depression, rheumatologic disease, COPD, heart disease, hypertension, prior pulmonary embolism or deep vein thrombosis, immunosuppression or chronic renal failure). Age was calculated per April 1st from the social security number. A child was defined as age below 16 years. Body Mass Index (BMI) was calculated from weight and height, and ISO-BMI [11] was used for children to enable comparison with adults. Participants aged 12 or older were asked about marital status (single, married, cohabitant, divorced or widowed, dichotomized to single or cohabitant), education level (primary school, secondary school, university or college), main occupation (student, worker, self-employed, retired, disabled or unemployed, re-categorized as student, worker or unemployed) and tobacco use (daily or sporadic, dichotomized to yes or no).
Hospitalization, vaccination, renal or renal complications
All hospitalizations related to the outbreak were reported and available to the research group, and hence the hospitalization rate was calculated both for the study population and the total population of Ulvik. Participants 12 years or older were asked and electronic patient records reviewed for both COVID-19 vaccination status and renal or vascular complications following infection (cerebral hemorrhage or infarction, transient ischemic attack, deep vein thrombosis, pulmonary embolism or renal failure) both at two- and seven-months follow-up.
Statistics
Descriptive data were presented as percentages, mean, median and interquartile range (IQR). Pearson’s c2 was used for comparison of proportions between groups with counts of at least 5 or above in each group, and Fisher’s exact for groups with less than 5. The Mann Whitney U-test was used for comparisons of continuous variables. Binary logistic regression was used to calculate odds ratios for binary variable outcomes, and were reported with 95% confidence intervals and p-values. For longitudinal analyses only participants who had answered both questionnaires were included. Statistical analyses were performed using IBM SPSS Statistics for Windows, version 28 (IBM Corp., Armonk, N.Y., USA). Figure 2 was made with SankeyMATIC (www.sankeymatic.com).