South Korea’s cancer screening proxies shifted sharply after COVID-19
South Korea’s national screening programs for solid cancers did not move through the pandemic as a single unit. The figures and tables attached to this analysis show program-specific screening-derived proxy rates changing across stomach, colorectal, breast, cervical, liver, and lung cancer, with the COVID-19 period dividing the timeline at 2020. The most important feature is not simply whether screening fell or rose; it is the unevenness of the rebound afterward. Some programs appear to have returned toward their prior trajectory, while others remained displaced enough to alter the longer arc of forecasted screening activity.
The analysis uses screening-derived proxy rates per 100,000 screened or eligible persons, which makes the data a measure of service activity rather than cancer incidence. That distinction matters clinically because screening volume functions as an early warning system for downstream delays in diagnosis. If fewer people enter a screening pipeline, fewer precursor lesions are detected, fewer incidental abnormalities are worked up, and more cancers risk being found later, when treatment is more complex and less forgiving.
COVID-19 did not simply interrupt appointments. It changed the denominator, the timing, and the rhythm of preventive care.
Program-specific trends were not interchangeable
The figure set makes clear that cross-program comparisons require restraint. Liver cancer and lung cancer sit on different denominators from the other programs, so a visual glance at the y-axis can mislead if the data are treated as directly comparable. Even within the same national screening system, each cancer program has its own entry criteria, testing logic, and practical barriers, which means one program can recover while another remains suppressed without any contradiction in the data.
The heatmap in the final figure captures this point neatly by standardizing annual proxy rates within each cancer program. Red and blue cells are not just color cues; they mark the internal oscillation of each program across the 2010–2024 span. Lung cancer begins only in 2019, so the pre-pandemic baseline is too short for a valid pre/post comparison in the same way as the other programs. That limitation is built into the structure of the dataset and has to be respected when interpreting the pandemic-era shift.
Breast, cervical, stomach, and colorectal screening likely matter most for public health planning because they represent repeated, system-level contact points where disruptions can accumulate into a broader diagnostic backlog. A temporary dip in one year may be recoverable. A slower return over several years is more consequential because it suggests the screening system has not fully reabsorbed deferred care.
Forecasts point to a staggered recovery curve
The OLS model-based forecasts extend the observed trends to 2028, and those projections are clinically useful precisely because they do not assume a uniform rebound. Instead, they translate the post-COVID trajectory into a forward-looking estimate of how screening-derived proxy rates may behave if current patterns persist. For stomach, colorectal, breast, cervical, and the composite measure excluding lung cancer, the forecast lines preserve the observed-period slope rather than forcing an artificial normalization.
In practice, that means the screening system may not simply snap back to its pre-2020 path. Forecasting by ordinary least squares is a straightforward statistical move, but it also carries a policy signal: if the underlying trend remains depressed or only partly restored, program managers should expect that deferred screening effects can persist well beyond the acute pandemic phase. The forecast intervals in Table 3 give the model uncertainty its due, yet the overall design still suggests a care environment in which recovery is uneven and program-dependent.
One implication is especially concrete. Screening programs are not just administrative exercises; they are the front end of cancer detection. When their activity falls, the effect is delayed, cumulative, and often invisible until later-stage presentations begin to surface.
Why the pandemic-era dip matters beyond 2020 itself
The pandemic likely affected screening through multiple mechanisms at once: reduced health-seeking behavior, deferred clinic visits, congestion in diagnostic services, and the redistribution of healthcare capacity toward urgent infection-related care. Even after services resume, patient behavior does not instantly normalize. Missed appointments can become missed years. In a screening context, a year of interruption is not equivalent to a year of delay in a low-stakes administrative process; it is a lost chance to detect disease before symptoms develop.
That is particularly relevant for cancers where screening intervals and age-based eligibility are central to program design. If participation lags, the resulting gap can alter not just the number of screens performed but the mix of lesions found, the timing of follow-up workups, and the burden placed on diagnostic pathways that were already strained during the pandemic. The proxy-rate framework cannot measure those downstream harms directly, but it does reveal where the system has not fully returned to baseline behavior.
Limitations are built into the measure, not hidden from it
This analysis relies on proxy rates derived from screening programs rather than individual patient-level cancer outcomes. As a result, it describes service utilization, not detection yield, stage migration, or mortality. It also cannot resolve why one program recovered faster than another because the figures and tables do not supply behavioral, operational, or regional breakdowns. Lung cancer is especially constrained by the short pre-COVID observation window, which limits the strength of any pre/post inference.
Even so, the overall pattern is hard to dismiss. South Korea’s cancer screening system experienced a pandemic-era perturbation that did not resolve evenly across programs, and the OLS forecasts suggest that some of those shifts may still be visible years later. The remaining questions are operational: which programs need the most targeted re-engagement, how quickly deferred participants can be brought back, and whether future screening demand will continue to run below the path it would have followed without COVID-19. Those answers will determine whether the observed gap is treated as a temporary interruption or as a durable change in preventive cancer care.