key: cord-0077549-ow9zqq1t
authors: Rannikko, Juha; Viskari, Hanna; Sirkeoja, Simo; Tamminen, Pekka; Kaila, Ville; Ala-Mursula, Anni; Huttunen, Reetta; Laine, Janne; Syrjänen, Jaana; Nuorti, Pekka J.
title: Population-based assessment of contact tracing operations for COVID-19 in Pirkanmaa hospital district, Finland
date: 2022-04-21
journal: Open Forum Infect Dis
DOI: 10.1093/ofid/ofac214
sha: d035562c94a7f3baa6350b4a9a508fe0db061ee6
doc_id: 77549
cord_uid: ow9zqq1t

BACKGROUND: The COVID-19 epidemic overwhelmed local contact tracing (CT) efforts in many countries. In Finland, SARS-Cov-2 incidence and mortality were among the lowest in Europe during 2020-2021. We evaluated CT efficiency, effectiveness, and transmission settings. METHODS: PCR test-positive COVID-19 cases and high-risk contacts in the population-based CT database of Pirkanmaa Hospital District (pop. 540,000) during June 2020–May 2021 were interviewed. RESULTS: Altogether 353,926 PCR tests yielded 4,739 (1.3%) confirmed cases (average 14-day case notification rate, 34 per 100,000 population); about 99% of confirmed cases and high-risk contacts were reached by contact tracing team. Of 26,881 high-risk contacts who were placed in quarantine, 2,275 subsequently tested positive (48% of new cases), 825 (17%) had been in quarantine ≥48 h before symptoms, and 3,469 (77%) of locally acquired cases were part of transmission chains with an identified setting. Highest secondary attack rates were seen in households (31%), health care patients (18%), and private functions (10%). Among the 311 hospitalized patients, COVID-19 diagnosis or exposure was known in 273 (88%) before emergency room admission (identified patients). Health care workers had the highest proportion of work-related infections (159 cases, 35%). The source of infection was classifiable in 65% and was most commonly a co-worker (64 cases, 62%). CONCLUSIONS: Our data demonstrates the role of effective testing and CT implementation during cluster phase of COVID-19 spread. Although half of newly diagnosed cases were already in quarantine, targeted public health measures were needed to control transmission. CT effectiveness during widespread community transmission should be assessed.

A C C E P T E D M A N U S C R I P T

Finland was one of the least affected European countries during the first two years of COVID-19 2 pandemic (situation November 30, 2021, 1) with no overwhelming of hospital capacity and one of the 3 smallest death tolls in Europe, even though governmental restrictions have been less stringent than in 4 many other European countries (2). This is likely due to several factors. Finland has the advantage of 5 having the lowest population density in the European Union (3), from European perspective lower than 6 average everyday person to person contacts (4), and above average trust between citizens and officials 7 (5,6). In addition, public health actions such as testing and contact tracing were implemented early in the 8 epidemic and coordinated nationally. 9

In May 2020, the Finnish government adopted "test, trace, isolate, and treat" as the national strategy (7). 10

However, isolations and quarantines had already been used locally in hospital districts in Finland since the 11 beginning of pandemic. Core contact tracing (CT) operations include effective testing, active case finding, 12 timely isolation of cases and quarantine of high-risk contacts to interrupt onward transmission. We Similarly with health care workers (HCWs); if they did not spend time together outside work, the 23 transmission was judged to be work-related. Events such as birthday parties, weddings and 24 housewarmings were classified as private parties, whereas shorter, less intensive, and less sustained free 25 A C C E P T E D M A N U S C R I P T time contacts were classified as meetings between friends and relatives. When the index case and 1 secondary case lived together full time or most of the time, the transmission was classified as a household 2 transmission. Stairwell transmission was defined to occur among residents who shared the same stairwell 3 in an apartment building and had no other known exposure within 10 days of first positive case. Only one 4 case per apartment was defined as stairwell transmission and secondary cases in each household were 5 defined as household transmission. 6

We defined CT efficiency based on the CT performance indicators developed by the ECDC (11): for 7 example, the proportion of cases where CT is initiated, proportion of high-risk contacts reached by phone 8 call, proportion newly diagnosed cases that were part of known transmission chains, proportion of 9 contacts who develop COVID-19 in quarantine (quarantined before positive test result). In addition, the 10 timeliness of initiating investigations and lack of delays in testing and contacting exposed persons are 11 important indicators of CT efficiency. By effectiveness we refer to assessing the CT system's effects on 12 the epidemiological parameters of the epidemic (e.g., reductions in proportion of positive tests, secondary 13 attack rates, the 14-day incidence rate), or other health outcomes (e.g., transmission chain interruption). 14

This is a retrospective analysis of our contact tracing registry data; because of acute public health 16 response, ethical review or informed consent was not required according to the Communicable Disease 17

Variants of concern and vaccinations 19

The sequencing method for VOCs has been described elsewhere (12) . In accordance with national 20 guidelines, the samples for sequencing were chosen randomly from positive PCR tests of clinical 21 laboratories and sent to Finnish Institute for Health and Welfare for sequencing. COVID-19 vaccinations 22

began on 27 th of December 2020 in Pirkanmaa HD. Vaccinations began from selected HCWs and long-23 term care facility residents and then continued to elderly with descending age groups (13). The majority of 24

vaccinated persons received Pfizer-BioNTech COVID-19 vaccine (Comirnaty) with a 12-week dose 1 interval. AstraZeneca (Vaxzevria, person >65 years) and Moderna (Spikevax) vaccines were used to a 2 lesser extent, both with a 12-week interval. 3

We used a Finnish computer software (SAI, Neotide Corp, Vaasa, Finland) There were few cases during June to August 2020 and the highest peak was in March 2021 ( Figure 1) . 16

During the seven months, defined as Fall 2020 (June 1, 2020 -December 31, 2020) there were less cases 17 (1618) than during the five months, defined as Spring 2021 (January 1, 2021 -May 31, 2021, 3121 cases) 18 (Table 1) . Median age was 33 years (inter quartile range 21-48) with lower median age in Spring 2021 19 than in Fall 2020 (31 vs. 36 years, respectively). CT was initiated in 100% of the laboratory confirmed 20 cases and 99.6% (4 719/4 739) were interviewed by telephone. The proportion of all high-risk contacts 21 reached is not known, but we estimate it to be over 90% among household members, friends, relatives, 22

workplaces, and schools and substantially lower in other places such as bars and night clubs. There were 23 248 high-risk contacts who were delegated to other hospital districts (no information whether they have 24

been reached) and 236 high-risk contacts (<1% of known high-risk contacts) who were not reached (or 1 quarantined). 2 3 4

Median time from symptom onset to the laboratory test was 2 days (inter quartile range (IQR) 1-3). The 6 mean processing time of laboratory test improved from 2.1 days in the Fall 2020 to 1.3 days in the Spring 7 2021 (no change in median times); the overall median processing time was 0 (IQR 0-0) days after the test 8 and 4 706 (99.3%) of index cases were contacted within 24h of positive laboratory results. Known high-9 risk contacts of the index cases were also contacted within 24h, but occasionally the list of contacts was 10 completed later than 24h after the test result. Altogether, the median time from onset of symptoms in the 11 primary case to the first contact tracing call was 3 (IQR 2-4) days. 12

Hospitalizations and deceased 13 Table 2 presents hospitalizations and COVID-19 associated deaths in our region. Of 4 739 cases, 311 14 (6.6%) required hospitalization of which 31 (10.0%) were treated in intensive care unit. The case-fatality 15 proportion within 28 days of the positive test result was 0.9% for all cases and 14% for hospitalized 16 patients. A larger proportion of COVID-19 positive cases required hospitalization during Fall 2020 than 17

Spring 2021 (9.0% vs. 5.3%, respectively). Among hospitalized cases, COVID-19 diagnosis was already 18 known in 166 (53%) and additional 107 (34%) were in quarantine or exposure to COVID-19 was 19 otherwise known before admission (identified patients); 38 (12%) patients were diagnosed with COVID-20 19 during hospitalization without prior evidence of COVID-19 disease or contact (unidentified patients). 21

The setting of transmission was identified for 3 469 (73%) newly diagnosed cases and were part of know 1 transmission chains. The setting was unknown for 1 042 (22%) cases (Table 3) . For 235 cases (5%), the 2 source of infection was abroad. In cases where the setting was known, the index case could be identified 3 in 2 057 cases (59%). In the rest, there were either many potential primary cases or the primary case could 4 not be determined. Most of the known transmission events occurred at home (Table 3) . Transmission 5 events detected at day care centers or schools accounted for 2.5% of all known infections (Table 3 and 4). 6

Thirty-four children contracted COVID-19 infection from day care (1.0%). Fifty-one infections of 7 children and teens under 18 years were traced to school (1.5% of all known sources or 8.4% of known 8 sources in age group of 7-18 years). The source of infection was another pupil in 40 (78%) of these and 9 the rest were from teachers. 10

Workplaces accounted for 13% of all known infections (458 cases) ( Table 3 and 4). Health care workers 12 constituted 35% (159 cases) of all work-related infections. In healthcare settings, most of classifiable 13 transmission occurred between co-workers (64 cases, 62%) or from unidentified (undiagnosed or not 14 known to be exposed) COVID-19 patients (34 cases, 33%). Only in 5 cases (5%) transmission occurred 15 from an identified (diagnosed or known exposed) COVID-19 patient. In 56 cases (35%) the source of 16 infection could not be classified. These were cases associated with epidemics in health care facilities with 17 many potential indexes (three different outbreaks with >10 identified cases). 18

We ordered 30 425 official quarantines for the known high-risk contacts. The total number includes 20 extensions of quarantine due to new positive household member and new quarantines for the same 21 individual in different setting. Quarantines were ordered to 26 881 individuals (Table 1) Table 3 ). The attack rate among workplace quarantines 1 was 4.3% (199/4 659). The proportion in schools was only 0.6% (38/5 959) and in day care 1.5 % 2 (31/1 986). 3

By May 31, 2021, 42.8% of Pirkanmaa HD residents had have received at least one dose of COVID-19 5 vaccination and 8.3% received two doses. Of those who had received one dose, 128 were later tested PCR 6 positive in Pirkanmaa HD; median time from first dose to symptom onset was 25 days. Fourteen cases had 7 received two doses before the positive result. Thirty-two of the vaccinated patients needed hospitalizations 8 (median age 57 years, 2 patients with two doses) and 6 died (median age 86 years, 1 patient with two 9 doses). 10

The first VOC (501.V2, Beta) was detected in a random sample taken January 4, 2021. First B.1.1.7 12 variant (Alpha) was detected January 10, 2021. In April 2021 the Beta variant was temporarily detected 13 from nearly half of the random samples sequenced in our region (n=394) (Supplement Figure) . By the end 14

May, over 80% of all sequenced samples have been the Alpha variant, but with B.1.617.2 (Delta) variant 15 rising in the last weeks of May (first detected from sample taken May 7, 2021). 16

During the study period, the average 14-day COVID-19 case notification rate (34 per 100 000 population) 18 in our region represented the median in Finland (14) but was substantially lower than in most European 19 countries (1). Of the laboratory-confirmed cases, almost 100% were reached and contacted within 24h 20 after the positive test result. The exact number of all high-risk contacts is unknown, but among household 21 members and other settings with known denominator, high proportion were contacted and provided 22

information. Of the high-risk contacts who developed laboratory confirmed disease, about half were in 23

quarantine before their positive test result. About four-fifths of all domestically acquired new cases were 1 part of known transmission chains. Importantly, almost 90% of hospitalized cases had been identified 2 (diagnosed or known to be exposed) already before admission to emergency room, reducing the risk of 3 nosocomial transmission. 4

Most published reports evaluating the effectiveness and efficiency of COVID-19 contact tracing have 5 been mathematical modelling studies and there are few population-based cohort studies with real-world 6 data (15-18). Systematically collected data on testing and contact tracing implementation in our region 7 suggested high efficiency and a contribution to reduction in onward transmission based on the CT 8 performance indicators developed by the ECDC (19). However, no national or international benchmarks 9 or goals are available for direct comparison. The limitations of quarantine in containing COVID-19 have 10 been noted in some reports (16). Because of its fast spread, SARS-CoV-2 control cannot rely solely on 11 contact tracing and quarantines, even in a country with low incidence. In our region, almost 27,000 12 individuals were placed in official quarantine as a result of contact tracing. Some 29% of new COVID-19 13 cases occurred in persons who were in quarantine before symptom onset and 17% were in quarantine 48h 14 before symptom onset, i.e. entire infectious period. Various factors influence the success of quarantines. 15

First, the transmission may have occurred abroad or in higher incidence regions within the country; 16 second, there may have been delays from symptom onset to contact call; third, the compliance with testing 17 recommendations might not have been consistent. Despite these potential limitations, contact tracing has 18 been an effective strategy in controlling the spread of SARS-CoV-2 in our setting, in contrast with some 19 previous reports from other countries (18, (20) (21) (22) . However, the overall epidemic control depends on, for 20 example, sufficient vaccination uptake and people's willingness and ability to reduce their exposure by 21 limiting contacts and maintaining physical distance. 22

Few cases (6.5%) required hospitalization and only 0.7% were admitted to intensive care. The overall 23 case-fatality ratio was also low and affected primarily the elderly as 79% of deaths were in persons >75 24

years old, similar to reports from other countries (23, 24) . Direct comparisons are difficult, however, as 1 testing thresholds, age distribution of cases, and hospital admission criteria may be different. 2

Contact tracing provided information on the transmission setting in 77% of confirmed cases in our region, 3 enabling public health officials to target restrictions for minimal harm and maximal effect. Household was 4 the most common setting of transmission, as noted before (25-28). The index case in the household 5 contracted COVID-19 most commonly from meetings with friends or relatives, workplaces, private 6 parties, or other contacts of known COVID-19 cases. Transmission setting data has uncertainties as it was 7 not possible to identify the index case in some of the places. Nevertheless, in most settings the index was 8 known, and the relatively low incidence in the region helped in identifying the probable place of 9 transmission. We also detected a few sporadic cases where the transmission likely occurred among 10 residents sharing the same stairwell in an apartment building without known contact such as discussing 11

with the index (72 cases, 45 different apartment buildings). In this setting the route of transmission is 12 unclear for us. 13

Although schools were open almost the entire study period, very few infections were traced to 14 symptomatic children at schools (51 cases, 1.5%) (Figure 1 ). Transmission in children occurred mostly in 15 households, and school transmission accounted only for 8.4% of known sources in school-age children. A 16 study from Germany also concluded that school-related origin of infection was unlikely in majority of 17 cases (29). The proportion of quarantined children who developed confirmed COVID-19 in our schools 18 was only 0.6% suggesting that the quarantines should be better targeted. However, there might have been 19 untested asymptomatic carriers. 20

Health care workers accounted for 35% of cases defined as work-related although they comprise only 8% 21 of the workforce in Finland. Some of the detected cases may be associated with the more common 22

practice of screening HCWs for asymptomatic virus carriage than in other workplaces, but the proportion 23 is still disproportionately high. The most common source for work-related HCW infection was a co-24 worker. Even though unidentified COVID-19 patients (undiagnosed or not known to be exposed) 25

accounted for only 12% of all hospitalized patients, they were associated with much higher proportion of 1 HCW infections than COVID-19 patients that were identified already before admission to emergency 2 room (30,31). A potential reason for this finding could be that the identified cases had been symptomatic 3 for a longer time and were therefore less infectious during hospitalization than the unidentified cases. The 4 different stage of the patient's infectious period is probably also one reason why no transmission to HCWs 5 in intensive care unit (ICU) was identified in our district; the ICU patients were not as infectious as the 6 they had been symptomatic for longer than COVID-19 patients not in ICU. The lower risk among ICU 7 workers has been noted before (32). 8

Unidentified COVID-19 patients have been noted to be a risk for their roommates in hospitals (33). 9

Patients who are exposed at the hospital and continued the hospitalization are especially problematic. 10 Some hospital outbreaks have been very difficult to control because the secondary cases had also been 11 hospitalized during their most infectious period. 12

Some 26 881 individuals were placed in quarantine during the study period and the median number of 13 quarantines per positive case was 2. The majority (92%) of contacts did not develop laboratory confirmed 14 COVID-19 during quarantine, a percentage similar to some other reports (25,34). During Fall 2020, 15 however, persons who were exposed to a confirmed COVID-19 case but remained asymptomatic were 16 rarely tested. The large number of quarantine restrictions of high-risk individuals may have helped to 17 avoid need for implementing restrictions impacting the whole population, such as curfew. Furthermore, 18 according to the Finnish Communicable Diseases Act, all persons ordered in quarantine or isolation are 19 eligible to receive an allowance for the duration of quarantine or isolation that is proportional to their 20 regular income. This financial support system is likely an important factor in the high compliance with 21 quarantine in Finland and containing the spread of the epidemic (35, 36) . 22

In the Summer 2021, Delta VOC spread throughout the world, including our region. It was estimated to be 23 twice as transmissible as previous variants (37). Therefore, the secondary attack rates in this study were 24 likely higher in Delta era. Due to the soaring case numbers, our contact tracing principles changed 25

considerably. Isolations continued mostly through digitalized process (text message with a request to fill 26 an online form) and only unvaccinated household members were placed in quarantine. Outbreaks in 27 health-care facilities were also a priority area for CT. Guidelines for seeking testing and quarantine 28 protocols changed frequently and database insertions was reduced to minimum, preventing collection of 29 reliable data. During this community spread phase, the contribution of CT to epidemic control decreased 30

in our region. Soon after the Omicron wave began, it became obvious that the volume of cases and 1 transmission speed exceeded CT capacity, and the rationale of CT operations needed reconsideration. 2 Therefore, CT was limited primarily to health care associated cases and national recommendations for the 3 general public emphasized the importance of self-isolation for individuals with respiratory symptoms. 4

In conclusion, our data provides evidence on the role of effectively implemented contact tracing 5 operations in maintaining epidemic control during COVID-19 cluster phase spread in our population. 6

Contact tracing provided data on major transmission settings and the high rate of isolations and 7 quarantines reduced onward transmissionNevertheless, even during a low incidence period, effective 8 COVID-19 control required a combination of public health measures, including societal restrictions. Study. Ann Intern Med. 2020/08/13. 2020 Dec 1;173 (11) Persons arriving from abroad were recommended to self-quarantine, but they were not ordered in official quarantine because of travel history. Above mentioned 18 cases were known to be exposed abroad. 

European Union -fourth update

1.351 or P.1: data from seven EU/EEA countries

European Center for Disease Prevention and Control. Overview of the implementation of COVID-6 19 vaccination strategies and deployment plans in the EU/EEA

COVID-19 country overviews: Finland. 10 2021

Effectiveness of non-pharmaceutical public health interventions against COVID-19: A systematic 13 review and meta-analysis. PLoS One

Quarantine 16 alone or in combination with other public health measures to control COVID-19: a rapid review

Front Public Heal

Universal Use of N95s in Healthcare Settings when Community 6

Covid-19 Rates are High

An outbreak 8 caused by the SARS-CoV-2 Delta variant (B.1.617.2) in a secondary care hospital in Finland

seroprevalence and asymptomatic viral carriage in healthcare workers: a cross-sectional study

The Risk of Severe Acute 14

Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) Transmission from Patients With 15

Undiagnosed Coronavirus Disease 2019 (COVID-19) to Roommates in a Large Academic Medical 16

Active Monitoring 18 of Persons Exposed to Patients with Confirmed COVID-19 -United States

Successful contact tracing 21 systems for COVID-19 rely on effective quarantine and isolation

Lessons from countries 24 implementing find, test, trace, isolation and support policies in the rapid response of the COVID-25 19 pandemic: a systematic review

We wish to thank the contact tracing unit of Tampere University Hospital for their 16