Estimates of Cases and Hospitalizations Averted by COVID-19 Case Investigation and Contact Tracing in 14 Health Jurisdictions in the United States

Abstract

Context: The implementation of case investigation and contact tracing (CICT) for controlling COVID-19 (caused by SARS-CoV-2 virus) has proven challenging due to varying levels of public acceptance and initially constrained resources, especially enough trained staff. Evaluating the impacts of CICT will aid efforts to improve such programs.

Objectives: Estimate the number of COVID-19 cases and hospitalizations averted by CICT and identify CICT processes that could improve overall effectiveness.

Design: We used data on the proportion of cases interviewed, contacts notified or monitored, and days from testing to case and contact notification from 14 jurisdictions to model the impact of CICT on cumulative case counts and hospitalizations over a 60-day period. Using the Centers for Disease Control and Prevention's COVIDTracer Advanced tool, we estimated a range of impacts by assuming either contacts would quarantine only if monitored or would do so upon notification of potential exposure. We also varied the observed program metrics to assess their relative influence.

Results: Performance by jurisdictions varied widely. Jurisdictions isolated between 12% and 86% of cases (including contacts that became cases) within 6 to 10 days after infection. We estimated that CICT-related reductions in transmission ranged from 0.4% to 32%. For every 100 remaining cases after other nonpharmaceutical interventions were implemented, CICT averted between 4 and 97 additional cases. Reducing time to case isolation by 1 day increased averted case estimates by up to 15 percentage points. Increasing the proportion of cases interviewed or contacts notified by 20 percentage points each resulted in at most 3 or 6 percentage point improvements in averted cases.

Conclusions: We estimated that CICT reduced the number of COVID-19 cases and hospitalizations among all jurisdictions studied. Reducing time to isolation produced the greatest improvements in impact of CICT.

FIGURE 1

Epidemic Curves Fitted to Observed Case Counts With Case Investigation and Contact Tracing Programs, and Estimated Cases Illustrating What Might Have Occurred Had the Programs Not Been Implemented^a Abbreviations: CICT, case investigation and contact tracing; other NPIs, other nonpharmaceutical interventions. ^aSolid lines are the epidemic curves at 4 evaluation locations fitted to their observed cumulative case counts with both CICT and other NPIs implemented (see Supplemental Digital Content Appendix Table A4, available at http://links.lww.com/JPHMP/A847). Dashed and dotted lines are the estimated curves illustrating the high (dashed) and low (dotted) estimates of cumulative total cases that might have occurred at these locations if CICT had not been implemented, and only other NPIs were implemented during the 60-day period. The differences between the solid and dashed or dotted lines show the benefits of CICT under low and high assumptions of program effectiveness, with greater divergence between the solid and broken lines indicating greater impact. All results assume that the effects of CICT and other NPIs were constant over the 60 days shown.

FIGURE 2

Effects of Improvements and Constraints to Case Investigation and Contact Tracing Performance Measures Compared With the Baseline Percent Cases Averted by the Programs^a,b Abbreviations: CICT, case investigation and contact tracing; other NPIs, other nonpharmaceutical interventions. ^aPercent cases averted by CICT calculated as percentage of total cases averted out of remaining cases after other NPIs were implemented. ^bResults shown assume notification is a sufficient trigger for contacts to quarantine (high effectiveness CICT). Results for our low CICT effectiveness scenario (ie, monitoring contacts is necessary for effective quarantine) are provided in Supplemental Digital Content Appendix Figure A3 (available at http://links.lww.com/JPHMP/A847). Baseline results are shown in Table 2. The proportion of Cases Interviewed and Contacts Notified were capped at 100% and 0% when the baseline percentage interviewed was greater than 80% or less than 20%. See Supplemental Digital Content Appendix Table A5a–b (available at http://links.lww.com/JPHMP/A847) for more detailed results.