Symptoms and risk factors for hospitalization of COVID-19 presented in primary care

ObjectiveTo extend knowledge of early symptoms as a precondition of early identification, and to gain understanding of associations between early symptoms and the development of a severe course of the disease. DesignRetrospective observational study SettingAustrian GP practices in the year 2020, patients above 18 years were included. ParticipantsWe recruited 22 practices who included altogether 295 participants with a positive SARS-CoV-2 test. Main outcome measuresData collection comprised basic demographic data, risk factors and the recording of symptoms at several points in time in the course of the illness. Descriptive analyses for possible associations between demographics and symptoms were conducted by means of cross table. Group differences (hospitalized yes/no) were assessed using Fishers exact test. The significance level was set to 0.05; due to the observational character of the study, no adjustment for multiplicity was performed. ResultsLittle more than one third of patients report symptoms generally understood to be typical for Covid-19. Most patients present with a variety of unspecific complaints. We found symptoms indicating complicated disease, depending on when they appear. The number of symptoms is likely to be a predictor for the need of hospital care. More than 50% of patients still experience symptoms 14 days after onset. ConclusionsUnderrating unspecific symptoms as possible indicators for SARS-CoV-2 infection harbours the danger of overlooking early disease. Monitoring patients during their illness using the indicators for severe disease we identified may help to identify patients who are likely to profit from early intervention. Data availability statementAll data referred to in the manuscript are available from: Department of General Medicine and Family Practice, Karl Landsteiner Privatuniversitaet, Krems, Austria Article SummaryO_ST_ABSStrengths and limitationsC_ST_ABSO_LIThis study investigates data on the course of COVID-19 collected exclusively from patients in primary care and explores a wide range of symptoms. C_LIO_LIGPs were free to make their own testing decision according to their clinical judgement, and they followed each patient individually from day 1 to day 10 or 14. C_LIO_LILimitations of our study concern the limited number of patients, due to the increased workload under difficult working conditions during the pandemic as well as the effort not being remunerated. However, the number of cases needed to identify group differences was calculated in advance, and this number has been reached. Our overall results are in accordance with our preliminary result analyses. C_LI

RT-PCR testing to detect a COVID-19 outbreak in Austria: rapid, accurate and early diagnosis in primary care (The REAP study)

BackgroundDelay in COVID-19 detection has led to a major pandemic. We report rapid early detection of SARS-CoV-2 by reverse transcriptase-polymerase chain reaction (RT-PCR), comparing it to the serostatus of convalescent infection, at an Austrian National Sentinel Surveillance Practice in an isolated ski-resort serving a population of 22,829 people. MethodsRetrospective dataset of all 73 patients presenting with mild to moderate flu-like symptoms to a sentinel practice in the ski-resort of Schladming-Dachstein, Austria, between 24 February and 03 April, 2020. We split the outbreak in two halves, by dividing the period from the first to the last case by two, to characterise the following three cohorts of patients with confirmed infection: people with reactive RT-PCR presenting during the first half (early acute infection) vs. those presenting in the second half (late acute), and people with non-reactive RT-PCR (late convalescent). For each cohort we report the number of cases detected, the accuracy of RT-PCR and the duration of symptoms. We also report multivariate regression of 15 clinical symptoms as covariates, comparing all people with convalescent infection to those with acute infection. FindingsAll 73 patients had SARS-CoV-2 RT-PCR testing. 22 patients were diagnosed with COVID-19, comprising: 8 patients presenting early acute, and 7 presenting late acute and 7 late convalescent respectively; 44 patients tested SARS-COV-2 negative, and 7 were excluded. RT-PCR sensitivity was high (100%) among acute presenters, but dropped to 50% in the second half of the outbreak; specificity was 100%. The mean duration of symptoms was 2 days (range 1-4) among early acute presenters, and 4.4 days (1-7) among late acute and 8 days (2-12) among late convalescent presenters respectively. Convalescent infection was only associated with loss of taste (ORs=6.02;p=0.047). Acute infection was associated with loss of taste (OR=571.72;p=0.029), nausea and vomiting (OR=370.11;p=0.018), breathlessness (OR=134.46;p=0.049), and myalgia (OR=121.82;p=0.032); but not loss of smell, fever or cough. InterpretationRT-PCR rapidly and reliably detects early COVID-19 among people presenting with viral illness and multiple symptoms in primary care, particularly during the early phase of an outbreak. RT-PCR testing in primary care should be prioritised for effective COVID-19 prevention and control. Research in ContextO_ST_ABSEvidence before this studyC_ST_ABSA comprehensive and effective test-trace-isolate (TTI) strategy is necessary to keep track of current and future COVID-19 infection in the UK and avoid a secondary wave later this year, as society reopens. As part of a wider TTI strategy, it is important to assess the feasibility of COVID-19 testing in primary care. We searched PubMed for implementation of SARS-CoV-2 testing in primary care using the following search terms: ("SARS-CoV-2" OR "COVID-19") AND "testing" AND ("primary care" OR "general practice"). We did not find any studies that met these criteria. Added value of this studyTo our knowledge, our study provides first evidence that extension of a National Influenza Surveillance Programme to include SARS-CoV-2 RT-PCR testing in primary care leads to viral detection among patients presenting with mild to moderate flu-like illness during a local outbreak of COVID-19. We show that the sensitivity of reverse transcriptase-polymerase chain reaction (RT-PCR), the technique to detect viral RNA, is high (100%) in the initial phase of the outbreak and among patients who were acutely unwell. Acute infection was associated with multiply symptoms: loss of taste, nausea and vomiting, breathlessness, myalgia and sore throat; but not loss of smell, fever or cough. We also show high correlation between reactive RT-PCR and seropositivity. Implications of all available evidenceOur findings suggest that RT-PCR can rapidly and reliably detect early COVID-19 among people presenting with viral illness and multiple symptoms in primary care, particularly during the early phase of an outbreak. Furthermore RT-PCR testing in primary care can effectively detect new COVID-19 clusters in the community and should be included in any strategy for prevention and control of the disease.