No abstract

Contact tracing is a non-pharmaceutical intervention (NPI) widely used in the control of the COVID-19 pandemic. Its effectiveness may depend on a number of factors including the proportion of contacts traced, delays in tracing, the mode of contact tracing (e.g. forward, backward or bidirectional contact training), the types of contacts who are traced (e.g. contacts of index cases or contacts of contacts of index cases), or the setting where contacts are traced (e.g. the household or the workplace). We performed a systematic review of the evidence regarding the comparative effectiveness of contact tracing interventions. 78 studies were included in the review, 12 observational (ten ecological studies, one retrospective cohort study and one pre-post study with two patient cohorts) and 66 mathematical modelling studies. Based on the results from six of the 12 observational studies, contact tracing can be effective at controlling COVID-19. Two high quality ecological studies showed the incremental effectiveness of adding digital contact tracing to manual contact tracing. One ecological study of intermediate quality showed that increases in contact tracing were associated with a drop in COVID-19 mortality, and a pre-post study of acceptable quality showed that prompt contact tracing of contacts of COVID-19 case clusters / symptomatic individuals led to a reduction in the reproduction number R. Within the seven observational studies exploring the effectiveness of contact tracing in the context of the implementation of other non-pharmaceutical interventions, contact tracing was found to have an effect on COVID-19 epidemic control in two studies and not in the remaining five studies. However, a limitation in many of these studies is the lack of description of the extent of implementation of contact tracing interventions. Based on the results from the mathematical modelling studies, we identified the following highly effective policies: (1) manual contact tracing with high tracing coverage and either medium-term immunity, highly efficacious isolation/quarantine and/ or physical distancing (2) hybrid manual and digital contact tracing with high app adoption with highly effective isolation/ quarantine and social distancing, (3) secondary contact tracing, (4) eliminating contact tracing delays, (5) bidirectional contact tracing, (6) contact tracing with high coverage in reopening educational institutions. We also highlighted the role of social distancing to enhance the effectiveness of some of these interventions in the context of 2020 lockdown reopening. While limited, the evidence from observational studies shows a role for manual and digital contact tracing in controlling the COVID-19 epidemic. More empirical studies accounting for the extent of contact tracing implementation are required.

Working papers of the Max Planck Institute for Demographic Research receive only limited review. Views or opinions expressed in working papers are attributable to the authors and do not necessarily reflect those of the Institute.

Indicators based a fixed “old” age threshold have been widely used for assessing socioeconomic disparities in mortality at older ages. Interpretation of long-term trends and determinants of these indicators is challenging because mortality above a fixed age that in the past would have reflected old age deaths is today mixing premature and old-age mortality. We propose the modal (i.e., most frequent) age at death, M, an indicator increasingly recognized in aging research, but which has been infrequently used for monitoring mortality disparities at older ages. We use mortality and population exposure data by occupational class over the 1971-2017 period from Finnish register data. The modal age and life expectancy indicators are estimated from mortality rates smoothed with penalized B-splines. Over the 1971-2017 period, occupational class disparities in life expectancy at 65 and 75 widened while disparities in M remained relatively stable. The proportion of the group surviving to the modal age was constant across time and occupational class. In contrast, the proportion surviving to age 65 and 75 has roughly doubled since 1971 and showed strong occupational class differences. Increasing socioeconomic disparities in mortality based on fixed old age thresholds may be a feature of changing selection dynamics in a context of overall declining mortality. Unlike life expectancy at a selected fixed old age, M compares individuals with similar survival chances over time and across occupational classes. This property makes trends and differentials in M easier to interpret in countries where old-age survival has improved significantly.

The U.S. elderly experience shorter lifespans and greater variability in age at death than their Canadian peers. In order to gain insight on the underlying factors responsible for the Canada-U.S. old-age mortality disparities, we propose a cause-of-death analysis. Accordingly, the objective of this paper is to compare levels and trends in cause-specific modal age at death (M) and standard deviation above the mode (SD(M +)) between Canada and the U.S. since the 1970s. We focus on six broad leading causes of death, namely cerebrovascular diseases, heart diseases, and four types of cancers. Country-specific M and SD(M +) estimates for each leading cause of death are calculated from P-spline smooth age-at-death distributions obtained from detailed population and cause-specific mortality data. Our results reveal similar levels and trends in M and SD(M +) for most causes in the two countries, except for breast cancer (females) and lung cancer (males), where differences are the most noticeable. In both of these instances, modal lifespans are shorter in the U.S. than in Canada and U.S. old-age mortality inequalities are greater. These differences are explained in part by the higher stratification along socioeconomic lines in the U.S. than in Canada regarding the adoption of health risk behaviours and access to medical services.