Embracing Controversy: A Second Look at CDC Reform Efforts in the Wake of COVID-19 Free

Abstract

The US Centers for Disease Control and Prevention (CDC) has responded to criticism claiming that the agency's COVID-19 response was lacking by proposing internal reforms intended to improve its performance during the next pandemic. The reforms are aimed at improving surveillance, analytic capacity, and agency communications. This article conducts a counterfactual analysis of the CDC's proposed reforms to ask how they might have changed outcomes in four cases of guidance controversy during the pandemic if they had been completed in advance of COVID-19. Although the CDC's planned reforms have merit, they are predicated on the ability to come to scientific closure in a highly charged political environment. To improve outcomes in a future pandemic, the agency should consider how it plans to communicate with the public when recovering from error and when addressing controversy spurred by criticism from credible experts. However, the ability of future presidents to limit CDC performance and communications in the next pandemic and the lack of political consensus around the value of independent public health expertise are likely to threaten any effort to improve pandemic response.

By now, Americans are familiar with the dismal statistics accompanying the US experience of the COVID-19 pandemic. Only months before the pandemic began, a 2019 report by the Global Health Security Index ranked the United States first among nations in terms of pandemic preparedness (Cameron, Nuzzo, and Bell 2019).¹ However, in terms of per capita morbidity and mortality, the United States fared worse than most other wealthy countries. Many critics blame the pandemic response of the US Centers for Disease Control and Prevention (CDC), focusing in particular on error-ridden and confusing guidance. Why should the public not disregard CDC guidance entirely if future guidance is likely to amend or overturn what the public is currently being told? When Dr. Rochelle Walensky, President Biden's first director of the CDC, received a draft report assessing the agency's COVID-19 response, she acknowledged agency failures and committed to a set of reforms (LaFraniere and Weiland 2022). Walensky promised to improve the information and analysis on which guidance is based as well as the agency's ability to communicate clearly with the public.

These CDC reforms aim to avoid error and controversy by improving surveillance, analysis, and the agency's ability to communicate the science that backs its guidance. Its portfolio of planned reforms, however, is incomplete. Responding to a pandemic driven by a novel infectious agent is an error-prone process. Whereas most science-informed policy making occurs years and even decades after the establishment of a stable scientific record, pandemic response often involves acting before critical learning has taken place. Moreover, even in domains where policy makers can rely on a stable evidence base often characterized by expert consensus, policy controversies can drag on for years (Collingridge and Reeve 1986; Jasanoff 1990; Kagan 1991; Landy, Roberts, and Thomas 1990; Mazur 1973; Nelkin 1979).

Reforms that aim to avoid error are insufficient; it is also crucial to recognize the inevitability of errors during a pandemic. Similarly, to the extent that agency reforms are geared to help the agency to avoid controversy, the agency is missing an opportunity to make use of controversy. Debate that can help separate strong claims from weak ones is a necessary part of learning during a pandemic. Reform plans that prepare the CDC to recover from error and embrace controversy would bolster the agency's reform strategies and contribute to improved performance in the next pandemic.

This article begins with an overview of the CDC's role during a pandemic. Next, it reviews criticisms of the CDC's performance during COVID-19 and introduces the agency's planned reforms. The article then examines four cases of controversy that emerged around the CDC's COVID-19 guidance and uses these to ask how well these reforms, had they been completed by December 2019, would have helped the agency during the pandemic. The analysis suggests that the CDC's reform plans are incomplete and that the agency should develop plans for how to recover from error and make use of controversies as a way to inform the public about the nature of learning when responding to the threat of a novel infectious disease.

The article does not consider how the CDC should contend with conspiracy theories and misinformation. Instead, it asks an earlier question. Specifically, during a pandemic driven by a novel infectious disease, can the CDC produce scientifically grounded guidance that might gain the support of expert stakeholders in a position to judge the quality of its expertise and policy efforts?

The CDC's Role during a Pandemic

The CDC's role in trying to protect public health is similar whether or not the country is confronting a pandemic driven by a novel infectious disease (Etheridge 2023; Henderson 2016). The agency's mission is designed to monitor, both nationally and globally, anything that might produce excess morbidity (illness or harm) or mortality (death) in the United States. To do this, it needs to be able to: (1) monitor trends in morbidity and mortality; (2) identify what is driving those trends; (3) aid in the design, dissemination, and implementation of strategies to prevent harm; and (4) monitor the effectiveness of those strategies to improve them over time (CDC 2024b; Henderson 2016). In these efforts, the CDC is a producer, funder, and consumer of scientific information that can inform prevention. It works with partners—typically state and local public health agencies as well as nongovernmental organizations and community-based organizations—to build local public health capacity and better understand its own performance in light of those partners’ experiences and insights (CDC 2024a, 2024b).

In its efforts to decrease morbidity and mortality, the CDC operates through the strength of the information it provides to the public health community. The agency, created in the 1940s, has a limited formal legal role in public health outside its jurisdiction at ports of entry (NASEM 2020). Instead, most public health authority is held by states, counties, and some cities (IOM 1988). Because of this federated public health structure, the CDC acts on public health only insofar as it can persuade stakeholders of the value of its advice. The agency has built a formidable reputation as a global leader in public health, although the CDC is not without its critics (Zeitvogel 2016).²

Other federal agencies that are expected to produce scientifically defensible policies have at least two advantages the CDC lacks when responding to a pandemic: (1) a relatively stable evidence base, and (2) time. For example, it is not uncommon for the Environmental Protection Agency (EPA) to spend five to 10 years in the process of promulgating new environmental health regulations (Landy, Roberts, and Thomas 1990). Even the EPA's process of producing just the preliminary scientific report that can be used as the basis for a new rule—its so-called integrated science assessment—takes the agency an average of more than four years.³

During its routine operations, the CDC's pace of incorporating science into policy is similar. A single monitor-diagnose-inform-respond cycle can take the CDC years to complete. In contrast, when addressing a pandemic driven by a novel infectious disease, that cycle is compressed into months or weeks. Such a rapid cycle leaves no room for the lengthy notice and comment process that provides opportunities for public deliberation and extended stakeholder engagement. These processes are not always successful; agency policy-making efforts often fail. When these processes do work, although they may fall short of silencing critics, they can produce policy outcomes that are able to withstand court challenges (Costa, Desmarais, and Hird 2016, 2019). Given the meager and rapidly evolving evidence base that emerges on the discovery of a novel infectious disease, the CDC must generate guidance in the absence of “scientific closure” (Hilgartner 2000), making pandemic response a particularly challenging performance environment.

Criticism of the US Response to COVID-19 and CDC Reform Efforts

Although a number of scholars argue that the poor performance of the United States during COVID-19 stems from factors outside the CDC's control (Arena et al. 2022; Béland et al. 2021; Cafruny 2023; Greer et al. 2023; Hacker et al. 2021; Huberfeld, Gordon, and Jones 2020; Lynch 2020; Singer et al. 2021; Zaki et al. 2022), many critics attribute multiple failures to the CDC itself. These include problems with the agency's case definition, its distribution of flawed tests,⁴ flawed guidance, guidance reversals, failures to communicate, and limitations in its capacity for data collection (table 1).

Acknowledging the criticism leveled at the agency, CDC Director Walensky launched an internal review of the CDC's performance during COVID-19. In August 2022, the reviewers completed a report of their findings titled “CDC Moving Forward Summary Report” (CDC 2024a). In it, reviewers criticize the agency for being too academic and too slow to properly respond to a pandemic. Reviewers also argue that the agency's siloed structure impeded core public health functions, and they recommended improved internal coordination and interaction with external partners to improve core public health functions (CDC 2024a). Noting that, before COVID-19, the agency managed relatively small outbreaks, the report suggested that the agency needs to increase its capacity for pandemic response by maintaining a larger group of staff trained for pandemic response and by having increased budget flexibility (CDC 2024a).

To address those shortcomings, the internal review listed five goals that should guide reform efforts: (1) share scientific findings and data faster; (2) translate science into easily understood policy; (3) prioritize communications; (4) develop a workforce prepared for emergencies; and (5) promote results-based partnerships. Walensky announced the agency's post–COVID-19 reforms, with an approach organized according to these five goals (CDC 2024a).

Some of the planned reforms are already underway. For instance, the CDC reports that its approach for developing and distributing tests was used successfully during Mpox (CDC 2024a).⁵ It has also initiated a novel electronic system—electronic case reporting, or eCR—for automatically collecting case data from providers (CDC n.d.). Another reform has been the agency's investment in the development of more sophisticated models for pandemic forecasting (CDC 2023c).⁶ There is also a promise to share real-time data. However, it is not clear that these reforms will allow the agency to share scientific findings more quickly than it did during the pandemic, since “scientific findings” and “data” might be distinct from each other.

Under the heading of communications, the CDC (2024a) is gathering feedback about how to improve the content and function of its website. To meet the goal of translating science into easily understood policy, the agency says it will present scientific language so anyone can understand it and is “standardizing guidance development” across the agency (CDC 2024a). To meet the workforce goal, the CDC is training its workforce for pandemic response and core public health functions and hiring new staff. It also indicates that it has created new tools to guide strategic deployment of staff in an emergency, but it does not describe those tools.

Another focus of the CDC's internal reforms is to improve its interactions with partners, including creation of a new center focused on gathering partner feedback. The agency also oversees a $3.7 billion public health infrastructure grant to improve workforce and infrastructure in more than 100 public health jurisdictions (CDC 2024a). In addition, it has reinstated an external expert advisory body to the director. The CDC has also added health equity to its reform efforts, which includes tracking health equity outcomes in case reporting, asking its advisory board to provide advice on health equity, and building local public health capacity (CDC 22024a).

These reforms promise to bring more and, hopefully, better data to the agency during a crisis. With increased modeling capacity, the CDC may be able to generate more reliable predictions about future states of an emerging outbreak. Both a better trained workforce within the CDC and more local public health capacity should allow the entire public health system to shift from routine operations to crisis operations more smoothly, and arguably to sustain that level of performance throughout a crisis. An external advisory board and more input from partners could increase the agency's insight during a pandemic.

However, the communications aspects of CDC reform efforts likely underestimate the problem of communications during a pandemic driven by a novel infectious disease. The agency recognizes the need to provide clear, accessible guidance; it aims to present the science underlying its guidance using language “anyone can understand” (CDC 2024a). This, along with its focus on standardizing guidance, suggests that the CDC sees guidance as a product. If it understands that these products must also be supported by ongoing communications processes, often with multiple stakeholders (Funtowicz and Ravetz 1993; Landy, Roberts, and Thomas 1990), that information is not clearly explained in its reform plans.

Much rests on the CDC's notion that it is easy to translate science into policy or that, if the agency has a strong scientific foundation underlying its guidance, there will be a simple way to explain the links between science and policy. A considerable body of social science literature establishes that controversy stems not necessarily from the quality of a policy's scientific inputs but from the distances between stakeholders’ preferred policy outcomes (e.g., Collingridge and Reeve 1986; Gieryn 1983; Jasanoff 1990; Landy, Roberts, and Thomas 1990; Mazur 1973; Nelkin 1979). The norm in US policy making when trying to produce science-informed policy is protracted debate (Jasanoff 1990; Kagan 1991; Landy, Roberts, and Thomas 1990). Given that pandemic response is highly politicized, reforms that do not explicitly include plans for engaging with anticipated controversy are incomplete.

The next section presents case studies treating four instances when CDC guidance became the subject of expert debate. A simple reading of the controversies surrounding these guidance failures is that the agency issued guidance outside of an existing evidence base and that better science would lead to better guidance. After a brief presentation of each case, the article asks whether better data, better analysis, or simplified communications would have changed the guidance and/or its reception during COVID-19.

Reviewing COVID-19 Guidance Controversies

During COVID-19, the CDC found itself enmeshed in controversies around its guidance. On two occasions it reversed its recommendations for masking, it struggled with political appointees to publish guidance that met its standards for science, and its updated isolation guidance at a critical point in the pandemic drew skepticism over its recommendation to use symptoms rather than a COVID-19 test to leave isolation. Even though these controversies may be familiar to anyone who lived in the United States during COVID-19, the following case descriptions provide a review of the events, the emerging science, and notable scientific uncertainties that shaped these controversies.

Case 1: Mistaken Starting Assumptions

The CDC's first guidance reversal came on April 3, 2020, when the agency recommended that the public use cloth face coverings (Dwyer and Aubrey 2020). Before this, the CDC had recommended against masking in public. Specifically, on February 12, 2020, in a press briefing, Dr. Nancy Messonnier, director of the CDC's National Center for Immunization and Respiratory Diseases, explained the agency's recommendations regarding masks to prevent COVID-19 transmission:

Messonnier was not the only official to carry the public health message about COVID-19 and masks. In a now-infamous tweet, US surgeon general Jerome Adams implored the public to stop buying masks, asserting that they were not effective in preventing the general public from contracting COVID-19 (Cramer and Sheikh 2020). Adams went on to explain that masks were needed by health care workers treating infected patients.

Both the CDC guidance and the surgeon general's tweet reflected the assumption that COVID-19’s spread would mirror that of SARS, where transmission occurs after a patient experiences fever, with most transmission occurring in the second week of illness (Apuzzo, Gebrekidan, and Kirkpatrick 2020; Keshta et al. 2021; Wilson 2020). If COVID-19 were like SARS, masking on the part of those without any signs of infection would do little to interrupt transmission. Moreover, the use of masks by those without symptoms would divert masks away from health care workers caring for the severely ill, who were most at risk of contracting the disease.

Even though researchers and clinicians began to document suspected instances of asymptomatic transmission at the end of January 2020 (Apuzzo, Gebrekidan, and Kirkpatrick 2020; Han et al. 2020; Holshue et al. 2020; Rothe et al. 2020; Wilson 2020), the World Health Organization continued to argue throughout the spring that most transmission was symptomatic (WHO 2020). By March, a growing number of documented cases of possible asymptomatic transmission led to concerns about higher rates of asymptomatic transmission (Bai et al. 2020; Tong et al. 2020; Wei et al. 2020). At the end of March, CDC Director Redfield indicated that as much as 25% of transmission was being driven by asymptomatic cases (Whitehead and Feibel 2020).⁷ Before issuing new masking guidance, the CDC requested approval from the White House Coronavirus Task Force (Sun and McGinley 2020) and shared publicly that the new masking guidance was under consideration (Achenbach, Sun, and McGinley 2020). At the time the CDC updated its masking guidance, the European Centers for Disease Control and Prevention, which provides public health guidance for EU countries, continued to recommend against masking in public (Dwyer and Aubrey 2020).

Case 2: Political Censorship

Following Nancy Messonnier's February 2020 press conference in which she shared the CDC's expectation that the disease could not be contained and asked the public to prepare for “significant disruption,” the Trump administration abruptly terminated regular public briefings from the CDC, with Trump himself taking over the role of providing public updates. Even though experts attended Trump's briefings, they did not prevent him from making unscientific and potentially dangerous claims (Clark 2020; Dawsey et al. 2020). In addition to direct communications from the CDC being terminated, journalists began reporting that guidance apparently published by the agency was being altered by White House officials (Weixel 2021). In one example, guidance noting the risk of COVID-19 transmission stemming from drinking from a shared communion cup was rewritten to remove any mention of COVID-19 risks related to participation in religious services (Valencia and Kelly 2020). Moreover, guidance addressing school reopening and when to test after COVID-19 exposure was altered by members of the White House Coronavirus Task Force (Roubein 2020; Wu 2020). The guidance documents stood out to external experts for their notable departures from prevailing public health standards and terminology, raising concerns that CDC guidance was being released without being subjected to the agency's normal scientific review (Mandavilli 2020).

Evidence also emerged that political appointees in the Department of Health and Human Services were successful in delaying the publication of data about COVID-19 in the CDC's Morbidity and Mortality Weekly Report (Diamond 2020). This raised additional concerns that the White House was interfering with the release of data about COVID-19 cases to try to claim that the pandemic was well under control. Finally, the White House directed hospitals to bypass a long-standing CDC data collection system and instead asked hospitals to report COVID-19 cases to an untested system managed by a private firm that would not make case data publicly available (Acosta and Cole 2020; Segers 2020; Stolberg 2020).

Case 3: Changing Conditions

From January through May 2021, COVID-19 vaccine went from being in short supply to being widely available. By May 1, 2021, 31% of Americans were fully vaccinated, with another 44% having received a first dose (USAFacts 2024). Studies assessing vaccine efficacy pointed to a number of promising findings. Several studies found low rates of breakthrough infections for those who were fully vaccinated, both from initial clinical trials (Baden et al. 2021; Polack et al. 2020) and from real-world studies (Chodick et al. 2021; Dagan et al. 2021). Researchers also found low viral loads among those infected after receiving only one dose of the vaccine (Levine-Tiefenbrun et al. 2021; Petter et al. 2021). Additional research found prevention of asymptomatic infection among those fully vaccinated (Jones et al. 2021; Tang et al. 2021). After reviewing the emerging evidence addressing vaccine efficacy, the CDC issued guidance on May 13, 2021, indicating that those who were fully vaccinated did not need to mask in public (Rabin, Mandavilli, and Weiland 2021).

The guidance prompted a range of criticism, including from experts who claimed the data supporting the guidance had not been sufficiently validated (Holtgrave and Rosenberg 2021) and from experts who argued the opposite—that the CDC had waited too long in the face of clear evidence to drop the masking requirement for those who were fully vaccinated (Shepardson 2021). Additional contemporaneous criticism focused on whether the guidance would be too difficult to implement or might encourage the unvaccinated to stop masking (Stanley-Becker et al. 2021).

By late July 2021, with evidence showing increased viral loads among breakthrough infections associated with the surging Delta variant, the CDC recommended that, in areas with high rates of transmission, even those who were fully vaccinated should mask in public (Howard 2021; Lovelace Jr., Repko, and Mendez 2021). Critics responding to the new guidance voiced several disparate concerns, including: (1) the evidence base had not really changed, as the CDC claimed (Stobbe 2021); (2) the CDC failed to be transparent in sharing data supporting the new guidance (Stolberg and Shear 2021); (3) the CDC was moving too slowly in updating guidance based on emerging evidence (Abutaleb et al. 2021; Aubrey 2021); and (4) the focus on breakthrough infections allowed public health messaging to stray from a focus on encouraging vaccination (Stolberg and Shear 2021).

Case 4: Generating Policy during Uncertainty

Responding to evidence that the less than a third of those who tested positive for COVID-19 were following the existing isolation period of 10 days, as well as research showing that most COVID-19 transmission happens in the first five days of infection (including two days before onset of symptoms), and anticipating a winter surge of COVID-19, the CDC issued new isolation guidelines in December 2021. The updated guidance relied on a combination of timing of infection and symptoms as a guide for when it was safe to leave isolation. Specifically, anyone meeting the following criteria could leave isolation after five days (CDC 2021, 2023b):

• ▪ mild and resolving symptoms;

• ▪ no fever for more than 24 hours;

• ▪ requirement to wear a well-fitting, high-quality mask when around others for a full 10 days after onset of symptoms.

For those experiencing a mild case of COVID-19 with no fever and resolving symptoms by day five, the CDC held that the risk of transmission was small and was sufficiently controlled by masking in public on days six through 10 (Bouton et al. 2022; Lefferts et al. 2022).

CDC Director Walensky explained the new isolation guidance in a series of media appearances (CDC 2022; Forgey 2021). Walensky argued that the previous 10-day guidance had failed to keep people focused on the most important period of transmission: two days before symptoms and two to three days after symptom onset (Forgey 2021; Weixel 2021). With an expected surge in winter cases, low levels of compliance with existing guidance, and an increasing number of people experiencing mild, even asymptomatic cases, agency officials predicted a large increase in transmission (Forgey 2021; Weixel 2021). Walensky argued that reinforcing the importance of isolation during the first five days, even for people who were asymptomatic, could help lower transmission during the expected winter wave (Collins, Bonifield, and Carvajal 2022; Forgey 2021).

In revising the isolation guidance, Walensky explained that the agency took into account what the agency “thought people would be able to tolerate” and said that a shortage of COVID-19 tests was not a factor in the new guidance (Weixel 2021). Walensky also explained that a positive COVID-19 test is not a reliable indicator of the ability to transmit COVID-19 (Forgey 2021; Weixel 2021). She clarified that the CDC expected the five-day isolation guidance would cut transmission rates when compared with the 10-day guidance. Referring to evidence that 85%–90% of transmission happens in the first five days of infection, the agency argued that masking during days six to 10, for those with resolving symptoms, would address the remaining chance of transmission (Weixel 2021).

Critics, including the American Medical Association, argued that a test was needed before people could safely end isolation (Carroll 2021; Harmon 2022; Wolters 2022). Some coverage of the guidance, while noting the controversy, focused on the CDC's defense of the guidance, essentially allowing the agency to provide an unfiltered version of its reasoning (Collins, Bonifield, and Carvajal 2022; Forgey 2021; Sun 2022; Weixel 2021). A number of news stories, on the other hand, argued that the CDC was ignoring clear evidence that people could still test positive after five days (Anthes 2022; McGregor 2022; Mueller 2021; Weissman 2021). The expert debate, however, was not about whether those who met the CDC's criteria could still test positive for COVID-19; many could. Instead, the debate turned on the more subtle question of whether those who met the CDC criteria for ending isolation were shedding viable virus (Cosimi et al. 2022). Both PCR and rapid antigen tests can produce positive results after an infected person is no longer contagious.⁸

To bolster the claim that the CDC guidance was out of step with existing research, one story cited three studies showing that around 50% of infected individuals could test positive between days six and 10 (Anthes 2022). None of the media-cited studies, however, tested the real question at the center of the guidance debate: what percentage of people with no fever and resolving symptoms who leave isolation after day five would be able to transmit COVID-19 while wearing a mask? The coverage (Anthes 2022) also failed to report disclosed conflicts of interests among study coauthors with financial stakes in companies producing COVID-19 tests (Hay et al. 2022).

Unlike the cases presented above, where the CDC admitted mistakes and reversed guidance, the CDC stood by its five-day isolation guidance. While its strongest critics portrayed the guidance as having no scientific merit and little expert backing, the policy was widely adopted by independent public health jurisdictions. Eleven independent public health authorities adopted similar guidance within a month of the CDC's recommendation.⁹ Moreover, several jurisdictions that initially required a test to leave isolation—Scotland, Japan, Australia, and California—aligned their guidance with the CDC's over the course of the next 15 months. The European Centre for Disease Prevention and Control (ECDC) split the difference in that it required a test to leave isolation except for cases of health care systems facing “extreme pressure” (ECDC 2022).

These cases provide a window into multiple challenges the CDC faced in trying to issue public health guidance for COVID-19. The focus within its reform efforts to increase its surveillance and analytic capacity suggests that the CDC, in reviewing its performance, believes that better insight into a novel infectious disease will improve the agency's performance. The next section, using counterfactual analysis, asks whether improved data collection, increased modeling capacity, and simplified communications would have changed outcomes in each of these cases.

Testing the Logic of the CDC's Planned Reforms

If the reforms the CDC outlined in its 2024 report (CDC 2024a) had been completed before December 2019, how might each of the cases presented above have played out? The CDC hopes to be able to generate better and faster insights about possible future states of a growing outbreak by increasing the efficiency and perhaps the scope of data collection through its electronic case reporting; improving the development, production, and distribution of reliable tests; and increasing analytic capacity through improved modeling. In case 1, where the CDC recommended masking in public after initially asking the public not to wear masks, better surveillance and analytic capacity could have reduced the time it took for the CDC to recognize the prevalence of asymptomatic transmission. The agency would need more than faster learning, however, to avoid its first COVID-19 guidance reversal.

Faster reduction of uncertainties does not guarantee there are no mistaken assumptions at the outset. The starting point for addressing a novel infectious disease is still one of profound uncertainty (Evans 2022). For example, no matter how efficient a country is in developing tests for a novel disease, the supply of available tests when that disease is discovered still starts out at zero. Moreover, faster data collection does not necessarily produce greater insight (Chiolero, Tancredi, and Ioannidis 2023). Data collection is profoundly shaped by case definition, which is in flux in the case of a novel infectious disease. For example, the case definition for COVID-19 was updated nine times—almost weekly—between December 2019 and March 3, 2020 (Tsang et al. 2020). Although more robust data collection and testing can reduce the time it takes to stabilize the scientific understanding of a novel infectious disease, they do nothing to change the starting point. Unless the CDC has a strategy for making no errors in its starting assumptions, case 1 should press the organization to consider how it plans to engage the public after it has issued erroneous guidance.

The idea that the CDC should plan for error may seem objectionable. If the CDC goes into a pandemic assuming that error is likely, does that mean that the CDC is never responsible for mistakes or ineffective policies? Alternatively, is there a reasonable performance standard that avoids this outcome without also requiring omniscience? Even though the CDC may lack sufficient evidence to know with certainty that a selected approach will work, the public can demand that the agency choose a defensible starting point and devote considerable resources toward testing its own assumptions so it is able to identify errors quickly and improve over time. This strategy will only work if elected officials allow for open, unadulterated communications between the agency and the public, which brings us to a distinct problem the CDC faced in case 2, where the Trump administration interrupted the content and normal channels of CDC communications.

Better surveillance and analytic capacity would have been unlikely to change the Trump administration's incentives to censor expert communications from the CDC. Moreover, the CDC's planned efforts to improve communications can function in a future pandemic only if the agency is allowed to communicate. In the United States, federal agencies are subject to a number of controls exercised by elected officials, including legislation (Guston 2000; Huber and Shipan 2002; McCubbins, Noll, and Weingast 1987; Moe 1989), oversight (McCubbins and Schwartz 1984), budgets (Wood and Waterman 1991), and management through the appointment of political appointees (Golden 2000; Wood and Waterman 1991). In addition, research demonstrates that a combination of fear and powerful norms to support elected officials can lead civil servants to self-censor when they disagree with political appointees (Golden 2000). Agencies in the past have relied on external allies, such as Representative Henry Waxman's efforts to better fund the CDC's response to AIDS in spite of stalling by the Reagan administration (Foreman 1994; Green 2011). They have also relied on strategic inaction, such as when the EPA removed a discussion of climate change from a report to prevent the insertion of unscientific language preferred by Bush administration officials (Revkin and Seelye 2003). Still, openly opposing political appointees is rare. This strategy is particularly risky during a pandemic, given that confronting democratically elected officials who enjoy political support can damage the agency's credibility with those same constituencies.

Case 3, where the CDC issued masking guidance for vaccinated people and then reversed it only months later, represents yet another challenge to planned reforms. Faster data collection, better testing, and increased modeling capacity will not guarantee that the agency will never issue guidance that must later be overturned because of novel disease dynamics that accompany new variants. One cannot, for example, collect data on variants that do not yet exist. More comprehensive and efficient data collection on breakthrough infections in the spring of 2021 could have produced a better estimate of the true rate of breakthrough infections. In this case, better data validation would still pertain only to the pre-Delta variants.

Should the agency simply plan to withhold guidance for a longer period of time to see if novel dynamics overtake the relevance of soon-to-be-released guidance? Practically speaking, there is no rule of thumb that can guide public health officials in deciding when to issue updated guidance. For every critic who argues that the CDC was moving too quickly (Holtgrave and Rosenberg 2021), there is a critic who lambasts the agency for being too slow and bureaucratic (Abutaleb et al. 2021; Aubrey 2021; CDC 2024a). This is not just a CDC problem. Experts addressing the outbreak in its first few weeks were not celebrated for trying to tease out whether they were seeing sustained versus unsustained human-to-human transmission (Boghani 2021). A methodical process of trying to validate evidence before issuing guidance can lead to allegations of a cover-up. The CDC's stated goal—to move from evidence to “easy to understand policy”—says nothing about how the agency should navigate such predictable sources of controversy.

Case 4, where CDC updated its COVID-19 isolation guidance, represents a subtle yet important challenge to the idea that planned reform efforts will change outcomes in a future pandemic. In this case, the lack of an a priori test of one's ability to transmit COVID-19 meant that public health officials had to rely on proxies to estimate transmissibility in the waning days of an infection. Planning to issue guidance based on solid evidence provides no insight into how the agency should approach issuing guidance when uncertainty is irreducible in the relevant time frame (Evans 2022; Funtowicz and Ravetz 1993; Weinberg 1972). Moreover, even when the science is stable and backed by a strong expert consensus, experts are still likely to disagree on remedies (e.g., geoengineering as a strategy to slow climate change).

The CDC's plan to be able to move from science to “easy to understand policy,” while a worthy goal, falls short of improving the agency's approach when policy recommendations are complex. If we trust antigen tests to indicate the onset of transmissibility, a full explanation of why they might be unreliable for identifying the end of transmissibility takes time and an attentive audience. In addition, many contextual factors shaped the CDC's isolation guidance. This only adds complexity to any effort to explain why the agency decided not to recommend a negative test before ending isolation. The 10-day isolation guidance was easy to understand. It was also ineffective and out of step with existing evidence. Thus, there may be situations when the public will need to learn something complex to understand the reasoning underlying public health guidance. When the ECDC issued its isolation guidance in January 2022, for example, it opted for complexity by creating 10 separate categories of guidance (ECDC 2022). Aiming to be clear and easy to understand is a laudable goal. The CDC should also develop strategies to communicate about complex matters of science and policy as part of its pandemic preparedness efforts.

Case 4 illuminates likely dynamics as experts within a professional community address uncertainty and context. Facing uncertainty, and in the absence of an obvious Pareto optimal solution, experts are likely to disagree (Funtowicz and Ravetz 1993; Pelling et al. 2022). Whereas some media coverage of the controversy framed it as a case of good versus bad guidance and informed (academic) experts versus uninformed (agency) experts, another framing, more consistent with the initial controversy and its evolution over time, is that public health experts generated two competing but defensible policy options.¹⁰ Both were informed; both had downsides. Importantly, both had credibility within the expert community, as demonstrated by their wide adoption. Especially important for the CDC is to take note of the dynamics that accompany its introduction of a policy that, over time, gained support across the public health professional community. If, in the short term, the CDC issuing what comes to be professionally accepted guidance faces the same dynamics as the CDC in a moment of failure, the organization has yet another reason to come up with a strategy that prepares the agency to make use of controversy rather than trying to avoid it.

Conclusion

The CDC's proposed reforms have an unmistakable logic to them. Who would argue against developing better surveillance and analytic capacity? At the same time, the CDC needs to be better prepared to lead in the absence of a time-tested evidence base and expert consensus. Necessarily, during a pandemic the agency will need to defend its policy recommendations based on evidence that is only months old, changing, and controversial.

Even though the organization identifies communications failures and aims to address them, these reform efforts are underspecified and appear to rest on the assumption that policies will flow naturally from a solid evidence base, rendering them easy to explain. To the extent that the organization presumes that better evidence will ameliorate its communications challenges, it underestimates the capacity for stakeholders in US policy making to argue about evidence. Other policy domains make it clear that having a stable expert consensus is not a panacea. But one need not exit the arena of pandemic response to identify shortcomings in CDC reform plans.

When learning about a novel infectious disease, error is the rule, not the exception. Moreover, external experts, whether or not they have superior information, are in a position to cast doubt on CDC policy. Since controversies among experts are rarely resolved in time frames that are relevant to pandemic response, the CDC may need to develop strategies to engage with critics more systematically. External critics, including stakeholders who lack scientific expertise, can point to limitations in CDC guidance and aid in efforts to address such limitations. When the CDC learns from its critics, it demonstrates responsiveness. In cases where its guidance withstands criticism, engagement with critics could result in a more informed public and growing confidence in openly debated reform strategies.

However, CDC reform efforts face another serious challenge that the agency has no effective means to address. The CDC may confront the next pandemic from a place of eroded public health capacity because it has not generated bipartisan support for increased personnel and greater resources (Devi 2020). Reports of political interference in data collection (Bergin and Keller 2021) and reporting (NPR 2020) during COVID-19 suggest that some elected officials prefer obfuscation to strengthening of public health systems. Challenges to public health authority have been especially strong in recent years among conservatives and Republican politicians (Weber and Achenbach 2023; Weber and Barry-Jester 2021). To that end, a number of states have cut state and local public health capacity and authority in response to COVID-19 (Sellers and Stanley-Becker 2021). If the agency is able to hire new staff and train its entire workforce in emergency response—goals included in its reform plans—the ability to make use of this expanded capacity is far from guaranteed. Civil service protections that have been in place for more than a century are newly under threat (Katz 2022; Knott and Miller 1987). Similarly, the CDC may be able to build systems for data collection and analysis. Less clear is whether those systems can be insulated from political attack. For example, how hard would it be for a future president to shut down the agency's electronic case reporting system? Even if these systems are operating as intended, plans for open communications with the public proceed at the pleasure of the sitting president.

From a public administration standpoint, there is little redundancy in our current national pandemic response system. While high-capacity state departments of public health could be a source of information and guidance were the CDC sidelined by elected officials, such departments cannot collect data outside of their jurisdictions. And as the COVID experience demonstrated, state public health agencies are also not immune to political intervention. Unfortunately for would-be reformers, the extraordinary partisan polarization in contemporary US politics makes it more difficult to forge a consensus on public health issues and could result in constrained authority for policy makers (Del Ponte, Gerber, and Patashnik 2024; Hopkins 2024).

Public health officials must lead pandemic response from whatever circumstances exist at the start of the crisis. The next pandemic might emerge when the CDC finds itself with less surveillance and analytic capacity and even less ability to communicate with the public than it had during COVID-19. Worse, it may face the next pandemic without civil service protections. For its part, the CDC needs to think much more concretely about how to inform the public amid controversy. As the CDC tries to increase its surveillance, analysis, and pandemic communications capacity, organizations outside the control of elected officials may also need to consider what role they can play in providing expert information to Americans during the next infectious disease crisis.

Acknowledgments

I would like to thank Frank Thompson and Heather Howard for suggesting this article and Andrea Campbell for recommending me to them. I want to thank reviewers, including Jonathan Oberlander, for their detailed and insightful comments. I especially want to thank infectious disease expert John Swartzberg for helping me to improve the presentation of the scientific debates underlying COVID-19 guidance.

Notes

References