Congruence of cancer screening recommendations between the USPSTF and the top ten US cancer centers: a cross sectional study

Summary

Background

The United States Preventive Services Task Force (USPSTF) is an independent volunteer organization of multi-disciplinary national experts in disease prevention that publishes evidence-based guidelines for cancer screening. However, these guidelines are not binding and many individual cancer centers publish their own screening recommendations.

Methods

This cross-sectional study compared the congruence of cancer screening recommendations between the USPSTF and the Top Ten US Cancer Centers as per the U.S News and World Report 2022–2023 for colorectal (CRC), lung (LC), cervical (CC), prostate (PC), and breast cancers (BC) in average risk patients. The evaluated variables included screening tests, recommendation direction, strength, frequency, age of start, age of end, and discussion/quantification of risks and benefits.

Findings

Our study identified substantial variability between the screening recommendations from the Top Ten US Cancer Centers and the evidence based USPSTF guidelines for BC, PC, and CC. The discordance was almost always in the direction of the cancer centers recommending more screening beyond the USPSTF's recommendations, and without discussion of potential risks and harms. For LC and CRC, the guidelines were generally concordant between the cancer centers and USPSTF.

Interpretation

Top Ten US Cancer Centers make screening recommendations inconsistent with those of the USPSTF recommendations, with cancer centers generally recommending more screening and omitting a nuanced discussion of benefits and risks. Given the national and international reputation of these cancer centers, their recommendations can have important population level implications, including confusion in the public.

Funding

None.

Research in context.

Evidence before this study

We searched PubMed and Google Scholar for studies published from database inception up to January 23, 2025, using combinations of the keywords “screening”, “cancer”, “cancer center”, “oncology”, “guidelines”, “U.S. Preventive Services Task Force”, “USPSTF”, and “overdiagnosis”. KW and BS screened abstracts for original research comparing screening guidelines and practices between U.S cancer centers and the U.S. Preventive Services Task Force (USPSTF) guidelines for colorectal (CRC), lung (LC), cervical (CC), prostate (PC), and breast cancers (BC) in average risk patients. Other types of articles, including reviews, commentaries, and studies evaluating other aspects of screening practices were used for background information. We identified four studies that were relevant to the comparison of screening guidelines between individual cancer centers and the USPSTF. Notably, these studies each focused on an individual tumour type, and included large numbers of cancer centers from various settings throughout the country. There is also substantial discussion in the literature surrounding overdiagnosis and overtreatment, and harm that this can cause to patients and the healthcare system alike. This can include financial toxicity, time toxicity, emotional distress, and physical harm. Several studies and reviews have discussed financial conflicts of interest at play given the substantial revenue associated with screening, and many cancer centers offer intensive screening packages that are not based on evidence.

Added value of this study

To our knowledge, no study to date has compared the USPSTF cancer screening guidelines to those of the top cancer centers in the country. Furthermore, no study has yet explored how this compares across several tumour types, nor have prior studies included all domains of screening, viz. eligible population, modality, frequency, age at start and stop, benefits, and harms. Our cross-sectional study provides a comprehensive comparison of the screening recommendations between the USPSTF and the Top Ten US Cancer Centers as per the USWNR for 2022–2023 for CRC, LC, CC, PC, and BC in average risk patients and highlights substantial discordance in this regard, especially for BC, PC, and CC with the direction of discordance almost always being towards the cancer centers recommending more screening than the USPSTF, without discussion of harms.

Implications of all the available evidence

Cancer screening is not a one-time event but a process that involves downstream consequences of further imaging, biopsies, surgeries, and treatment. USPSTF provides an independent, non-conflicted, evidence-based guidelines for cancer screening for US patients. Top rated cancer centers in the US have influence on practice patterns nationally, and even internationally. Results from our study, in combination with other existing literature, caution policymakers that the top-rated US cancer centers could have strayed away from USPSTF guidelines, almost always recommending more screening. The reasons for these discrepancies must be addressed, and a uniform policy for population level cancer screening should be instituted. At an individual clinician and patient level, these trends must be considered when making screening decisions, irrespective of the reputation of the cancer center making the recommendations.

Introduction

Cancer is the second leading cause of death in the United States, with lung, breast, colorectal, and prostate being the most common cancers contributing to this mortality. Interestingly, these are also the cancers central to population-level prevention efforts through screening and early detection and treatment.¹ The American Cancer Society (ACS) has recently expressed concern about the rising incidence of such “screenable” cancers.²

On the other hand, cancer screening may be associated with harms, specifically, overdiagnosis and overtreatment through detection of benign lesions that nevertheless need characterization and diagnosis via unnecessary and sometimes harmful procedures, or detection of slow growing incidentalomas that may never become symptomatic or hasten natural death.³ In addition, the emotional costs to the individual and financial costs to the system incurred during the screening and diagnostic investigations should be balanced against the potential benefits of cancer screening at the population level.

The United States Preventive Services Task Force (USPSTF) is an independent volunteer organization of multi-disciplinary national experts in disease prevention and evidence-based medicine.⁴ Created in 1984, this task force requires members to have no substantial conflicts of interest that could affect the scientific integrity of the recommendations. USPSTF publishes evidence-based recommendations on cancer screening, alongside various other conditions, solely based on scientific evidence.⁴ Although their recommendations are not binding, Grade A and B recommendations must be covered for by the insurance program under the Affordable Care Act.⁴ Thus, despite the presence of several screening guidelines in the U.S through professional societies, USPSTF recommendations hold a special place as a governmental recommendation created by members without conflicts of interest.

Since 1990, the U.S News and World Report (USNWR) has published an annual list ranking the U.S “Best Hospitals”.⁵ The USNWR ranking is a highly regarded list, which can act as a strong promotional tool for cancer centers looking to attract new patients. Highly ranked cancer centers are viewed as trusted authorities on cancer care, and, their websites are often looked to as credible sources for cancer related information, including screening. Cancer screening recommendations made on the websites of these cancer centers may impact screening practice not only within that cancer center, but also for the general public in and outside of the U.S., who may visit these websites looking for guidance. Given the wide-reaching influence these cancer centers have, it is crucial to assess whether the screening recommendations posted on their websites align with those issued by the USPSTF. Discrepancies between the two could lead to confusion for providers and patients alike, potentially affecting decision-making and patient outcomes. Thus, this study aims to evaluate the concordance between cancer screening recommendations from the USWNR's Top Ten US Cancer Centers with those issued by the USPSTF.

Methods

This is a cross-sectional study to assess the congruence of cancer screening recommendations and information between the Top Ten US Cancer Centers and the USPSTF guidelines.

Ethics

All the information presented in this study are publicly available and do not include patient data, and therefore, this study was not subject to ethical approval.

Selection of cancer centers

The list of Top Ten US Cancer Centers was obtained from the USNWR website (https://health.usnews.com/best-hospitals/rankings/cancer) for the 2022–2023 year on October 25, 2022. In order, these included University of Texas MD Anderson Cancer Centre (MDA), Mayo Clinic Comprehensive Cancer Centre (MC), Memorial-Sloan Kettering Cancer Center (MSKCC), Dana-Farber/Brigham and Women's Cancer Center (DF/BW), UCLA Jonsson Comprehensive Cancer Center (UCLA), Cleveland Clinic Cancer Center (CC), City of Hope Comprehensive Cancer Centre (COH), Hospitals of the University of Pennsylvania-Penn Presbyterian (HUP/PPMC), Northwestern Memorial Hospital (NMH), and Siteman/Barnes-Jewish Hospital (SCC) (Supplementary Table S1).⁶

Selection of cancer screening guidelines

We selected colorectal, lung, cervical, prostate, and breast cancers for our study because these are the most common “screenable” cancers accounting for high incidence and mortality. Screening recommendations for these cancers were searched for and extracted from the USPSTF website (https://www.uspreventiveservicestaskforce.org/uspstf) initially in October 2023, and most recently updated January 23, 2025. We focused on screening guidelines for an average risk population; recommendations for individuals at a higher risk of cancer (e.g. due to family history or genetic syndromes) were excluded. Categories of USPSTF recommendations range from A to D and I where A and B recommend offering screening, C suggests offering only for selected patients based on individual circumstances, D recommends against screening, and I denotes insufficient evidence. We then searched the websites of the Top Ten US Cancer Centers for screening recommendations for these cancers for the average-risk population.

Data extraction

Initial data extraction was initially conducted October–November 2023, and last updated January 2025. One author (KW) extracted the initial data from the guidelines and cancer center websites; data was verified by a second author (BS). Any discrepancies were settled by discussion with the senior author (BG). Cancer center websites were thoroughly searched for any published webpages containing cancer screening recommendations, including both the landing pages and any linked pages that could contain relevant information. If conflicting screening information was found on the same cancer center's multiple websites, the most recently updated page or the page most congruent with USPSTF were used for the purposes of data extraction. The full list of cancer center webpages with screening information is included in the Supplementary Appendix.

Assessment of congruence

The congruence between the recommendations from USPSTF and Top Ten US Cancer Centers for the selected five cancer sites were compared with regards to the following domains: modality for screening, recommendation strength and direction (for/against), frequency (e.g.: every year), age of start (e.g.: starting at age 40), and age of end (e.g.: stop at 80). Lack of congruence in any of these domains were considered as overall incongruent with USPSTF guidelines. In addition, we also extracted information on whether the websites provided any discussion of potential risks and benefits of screening. In order to classify as discussing potential risks of screening, cancer centers had to at least acknowledge the risk of false positive or negative results leading to further tests, potential side effects and/or complications of tests. The results are presented descriptively. Congruence was also characterized into one of the seven categories via color coding for easy visualization as described in Supplementary Fig. S1.

Role of funding source

There was no funding source for this study.

Results

Fig. 1, Fig. 2, Fig. 3, Fig. 4, Fig. 5 compare the cancer screening recommendations for each Top Ten US Cancer Centers to those of USPSTF for lung, colorectal, breast, cervical, and prostate cancers, respectively. Supplementary Fig. S1 provides a legend for interpretation of color codes.

Fig. 1.

Lung cancer screening recommendations in average risk patients from United States Preventive Services Task Force (USPSTF) and the Top Ten US Cancer Centers.

Fig. 2.

Colorectal cancer screening recommendations in average risk patients from USPSTF and the Top Ten US Cancer Centers.

Fig. 3.

Breast cancer screening recommendations in average risk patients from USPSTF and the Top Ten US Cancer Centers.

Fig. 4.

Cervical cancer screening recommendations in average risk patients from USPSTF and the Top Ten US Cancer Centers.

Fig. 5.

Prostate cancer screening recommendations in average risk patients from USPSTF and the Top Ten US Cancer Centers.

Lung cancer

The screening guidelines for lung cancer had the highest congruence between the cancer centers and the USPSTF (Fig. 1). The 2021 USPSTF guidelines recommend screening with an annual low-dose CT chest (grade B) adults between 50 and 80 years of age with a ≥20 pack-year smoking history who are current smokers, or quit within the last 15 years.⁷

Screening recommendations across all cancer centers were all generally in line with this, and some centers directly acknowledged the USPSTF guidelines (MSKCC, HUP/PPMC, SCC, DF/BW, NMH) but no cancer center provided a strength of recommendation. MC did not provide an upper age cut off to their screening recommendations. CC had two sets of conflicting screening recommendations; one was in line with the USPSTF, while the other recommended screening for those between 55 and 77 years old, and at least a 30-pack year smoking history. NMH, MC, SCC, COH, and CC acknowledged potential harms of lung cancer screening and only NMH included a discussion about the expected mortality gains from screening. No center discussed number needed to screen (NNS) or number needed to harm (NNH).

Colorectal cancer

For colorectal cancer, the 2021 USPSTF guidelines recommend screening adults (grade B for adults 45–49 years age, grade A for adults 50–75 years age) with one of: a colonoscopy every ten years, a flexible sigmoidoscopy or CT colonography every five years, a FOBT or FIT every year, sDNA FIT every one to three years, or the combination of a flexible sigmoidoscopy every ten years with an annual FIT (Fig. 2).⁸ There was also a weak grade C recommendation that screening may be selectively offered to adults aged 76–85 years on a case-by-case basis with consideration of their overall health, prior screening history, and preferences. Screening is not recommended for adults over 86.

Screening recommendations across the Top Ten US Cancer Centers were relatively congruent with these recommendations, however often incomplete. Only the UCLA and COH recommendations completely matched those of USPSTF. COH mentioned both USPSTF and ACS screening guidelines on their webpage. DF/BW also referred to USPSTF in their webpages, though did not include flexible sigmoidoscopy plus FIT as a potential screening tool and did not specify a recommended frequency for flexible sigmoidoscopy or CT colonography. NMH referred to USPSTF on their webpage, however also listed barium enema and digital rectal exam (DRE) as potential screening tools. Other centers recommended screening with some diagnostic tests and frequencies, however, did not include all test options compared to USPSTF (NMH, MDA, MC, HUP/PPMC, SCC). MSKCC did not specify a screening frequency for any of the listed diagnostic tests. CC advised a colonoscopy every ten years, or the combination of a yearly stool test and flexible sigmoidoscopy every five years, which was increased frequency compared to USPSTF. CC also mentioned CT colonography, flexible sigmoidoscopy alone, and stool-based tests alone as screening options, but did not specify a frequency. HUP/PPMC, CC, NMH, and SCC did not specify an upper age cut off to stop screening. None of the cancer centers directly provided strength of recommendation. MC, MDA, DF/BW, and UCLA acknowledged potential harms and only DF/BW discussed expected survival gains from CRC screening. No center discussed NNS or NNH.

Breast cancer

Breast cancer screening recommendations had substantial variability across cancer centers (Fig. 3). The 2024 USPSTF guidelines recommend screening for women aged 40–74 every 2 years (grade B) with a conventional digital mammogram (DM) ± digital breast tomosynthesis (DBT) and conclude insufficient evidence to recommend screening in women aged ≥75 years.⁹

Every Top Ten US Cancer Centers recommended increased screening frequency compared to USPSTF. CC and MSKCC recommended women complete self breast examinations starting at 20 years old. SCC recommended self breast examinations starting at age 25. Clinical breast examinations were recommended by CC starting at age 40, by MDA, MSKCC, and SCC at age 25, and by DF/BW at age 18. COH and UCLA were found to have multiple sets of conflicting recommendations. One COH site advised annual DM, while the other advised that DM or DBT can be completed every 1–2 years after age 55. Similarly, one UCLA site advised annual DM or DBT, while the other recommended DM every 1–2 years. Other cancer centers recommended DM or DBT at an annual frequency, except DF/BW and HUP/PPMC who also advised DM every 2 years could be considered after the age of 55, and CC also recommended that DM could be decreased to biannual for women aged 40–49 “if appropriate”. The only incidence of decreased screening was that DF/BW did not recommend mammograms until the age of 45. Finally, only COH and HUP/PPMC acknowledged stopping screening once a woman's life expectancy is 10 years or less, but none specify an upper age cut-off. None of the cancer centers provided strength or grading of their recommendations. Only CC included a discussion about the expected survival gains, and only MDA, MC and UCLA discussed potential harms from breast cancer screening; DF/BW advised patients to discuss risks and benefits of screening with their healthcare provider. No center discussed NNS or NNH.

Cervical cancer

The 2024 USPSTF draft guidelines for cervical cancer recommend screening every three years with cervical cytology alone in women aged 21–29 years, then human papillomavirus (hrHPV) testing alone every five years in women aged 30–65 (grade A) (Fig. 4).¹⁰ Alternatively, women aged 30–65 years may continue cervical cytology alone every 3 years or co-testing every five years with hrHPV and cervical cytology (grade A). Screening was not recommended for average risk women aged less than 21, or greater than 65 with adequate prior screening, or those with hysterectomy and removal of the cervix who do not have a history of a high-grade precancerous lesion (grade D).

The recommendations for MSKCC, and DF/BW were in line with these recommendations. COH was found to have two sets of conflicting screening recommendations; one was in line with those of USPSTF, while the other did not include hrHPV testing alone every 5 years as an option. DF/BW directly referred to USPSTF guidelines. HUP/PPMC cited ACS guidelines and recommended screening after age 25 with hrHPV testing every five years, and listed cervical cytology every three years or co-testing every 5 years as alternative options. MDA recommended cervical cytology every 3 years, or HPV test every 5 years beginning at age 25, followed by HPV or co-testing every 5 years, or cervical cytology every 3 years from ages 30 to 64. MC advised cervical cytology starting at age 21, and that co-testing could be initiated at age 25. Otherwise, co-testing or HPV alone every 5 years. The recommendations from CC were the most inconsistent, recommending annual cervical cytology for three consecutive years starting at the age of 18 or when sexually active, then every three years thereafter. They also did not advise screening for women aged 60 or greater. CC, NMH, UCLA, and SCC did not include hrHPV testing alone every 5 years as an option. CC also did not include hrHPV combined with cervical cytology every 5 years as an option for women aged 30–65. SCC did not specify an upper age cut off to discontinue screening, and MSKCC, DF/BW, HUP/PPMC, or SCC did not specify screening could be discontinued after hysterectomy with removal of the cervix. None of the cancer centers provided a strength of recommendation. Only CC discussed the potential harms of screening for cervical cancer. None of the cancer centers quantified the expected mortality gains, NNS or NNH.

Prostate cancer

The 2018 USPSTF guidelines advised that periodic prostate cancer screening with prostate-specific antigen (PSA) can be considered for men aged 55–69 on an individual basis, after discussing the potential harms and benefits of the test and incorporating their values and preferences into the decision (grade C) (Fig. 5).¹¹ Screening is not recommended for men aged 70 or greater (grade D).

All cancer center recommendations varied from those of USPSTF for prostate cancer. UCLA referred directly to the American Urological Association (AUA) guidelines, which advised baseline PSA ± digital rectal exam (DRE) between the age of 45 and 50 years, then every two to four years up to 69 years as shared decision making after discussing risks and benefits. MC advised PSA screening could be started after the age of 50 until 70 after discussing risks and benefits. SCC advised PSA testing between the ages of 45 and 75 after discussing potential risks and benefits. MDA advised PSA screening for men aged 45–84 years after a discussion about risks and benefits of screening. They noted that patients should discuss if screening is still appropriate after the age of 75. MSKCC also advised PSA screening for men over the age of 45. After the age of 70, they advised PSA should only be checked if a past test was high, and over the age of 76, noted that further testing is rarely helpful. DF/BW and NMH recommended discussing PSA screening with a physician at age 50 for average risk men, and did not provide an age cut off to discontinue screening. CC had two conflicting recommendations on their websites-one advised PSA starting at age 50 and the other recommended starting at age 45, with no cut off age. COH cited both ACS and USPSTF guidelines on their webpage. All centers except MSKCC also recommended a DRE to be considered as part of screening. All centers except HUP/PPMC and SCC included a discussion about potential harms of prostate cancer screening. No cancer center provided a strength of screening recommendation, quantified the expected mortality gains, or provided NNS or NNH.

Discussion

In this study, we found that the Top Ten US Cancer Centers did not necessarily follow the USPSTF screening recommendations in terms of promotion or patient education through their websites. Although there was some concordance between the screening recommendations from the USPSTF and the cancer centers for lung and colorectal cancers, we discovered substantial differences in recommendations for breast and prostate cancer, and to a lesser degree, cervical cancer. The direction of discordance was usually towards the cancer centers recommending more screening than the USPSTF. Given that these top-ranked centers have influence in academic, clinical, and public engagement settings, as well as international reputation, their recommendations have important implications for policy and implementation.

Top ranked cancer centers often promoted more screening beyond the USPSTF recommendations, which involved either more modalities, higher frequencies, younger starting age, and/or no or older stopping age. For example, all centers advised yearly mammograms for breast cancer screening, which is in contrast with the USPSTF guidelines that recommended biannual screening. Additionally, although evidence shows lack of benefit in screening women over 75 for breast cancer, no specific cut off age was specified for 9 of the Top Ten US Cancer Centers.⁹ Similarly, the USPSTF did not recommend prostate cancer screening in men over the age of 70, however, half of the centers offered screening past this age. The USPSTF recommended a very weak grade C for the initiation of PSA testing for prostate cancer beginning at the age of 55 through an individual decision based on values and risk factors; however 8 cancer centers encouraged consideration of screening earlier than this age for average-risk men, at either 45 or 50 years old, and none of them mentioned that the recommendation was weak. Additionally, one center recommended starting cervical cancer screening with cervical cytology starting at age 18, compared to age 21 per the USPSTF. It is noteworthy that divergence was highest for prostate cancer and breast cancer–the two cancers that are most prone to the harms of overdiagnosis and overtreatment.¹²

Furthermore, several cancer centers recommended intrusive clinical exams as screening tools, including DRE, and self or clinical breast examinations despite poor evidence to support their use. In addition to PSA testing, DREs were recommended for prostate cancer screening by 9 cancer centers. A DRE was also listed as a potential screening tool for colorectal cancer by one center. In women, clinical and self breast examinations were recommended starting as early as 18 years old. However, DREs, and breast examinations are no longer recommended as screening methods for cancer by many medical organizations and experts, including the USPSTF.¹³

The websites of top cancer centers seem to disproportionately highlight the benefits of screening without adequately discussing the harms (Fig. 6). Studies have suggested that the public overestimates the benefits and underestimates the harms of cancer screening.¹⁴ Our study shows that information on cancer centers’ websites can perpetuate these misperceptions. While all cancer centers discussed benefits of screening, only eight discussed the harms of screening for prostate cancer, five for lung cancer, four for colon cancer, three for breast cancer, and one for cervical cancer. Of note, several cancer centers recommended discussing potential risks of screening with a health care provider (HCP), the possibility of “slight discomfort”, or the need for further tests. This was not counted as an adequate discussion of potential harms for the purposes of this study. Furthermore, there was very limited quantification of mortality gains from screening for any cancer. No cancer center discussed number needed to screen (NNS) or number needed to harm (NNH) for any modality for any cancer. In addition to the physical and emotional toll on patients, overscreening and overdiagnosis of asymptomatic cancer is also associated with financial and time toxicity to individual patients and the health care system in general.¹⁵ Under the Affordable Care Act, any USPSTF recommendations with an A or B rating must be covered without cost sharing by Medicare and private insurers.¹⁶ When cancer centers deviate from these recommended practices, there may be hidden costs borne by individuals out of pocket. This discordance and lack of discussion of harms or the NNT or NNH has the potential to mislead the public and promote overdiagnosis and overtreatment, as well as confuse the public and erode their trust in cancer screening thereby discouraging adherence to even the USPSTF recommended cancer screening altogether. At a minimum, we recommend that all screening recommendations to the public must contain modality, age of start, age to stop, frequency, risks and benefits, as well as NNS and NNH. In addition, other subtle harms of more screening such as financial toxicity, time toxicity, anxiety, etc. are not captured by NNS or NNH.

Fig. 6.

Cancer centers whose websites included a discussion of the potential harms associated with screening, stratified by cancer type.

Furthermore, there is a lack of clarity regarding the development process of the recommendations from each cancer center. While some centers directly referenced the USPSTF or other professional society guidelines, many did not. It is likely that the consensus of treating providers at the cancer centers played a significant role in the development of these recommendations; however, the extent of this involvement is not clearly described. Indeed, some of the discrepancies between the Top Ten US Cancer Centers and the USPSTF can be explained by guidelines from other bodies with conflicting recommendations. Although guidelines from both professional societies and the USPSTF are based on available evidence, they are reviewed and updated on different timelines. For example, the American College of Radiology and National Comprehensive Cancer Network (NCCN) advise annual mammography starting at age 40.^17,18 NCCN further advises a clinical breast examination every 1–3 years from age 25 to 40, and annually afterwards. The American College of Obstetricians and Gynecologists (ACOG) advises mammograms every 1–2 years starting at age 40, with consideration of biannual mammograms after age 55.¹⁹ ACS advises that mammograms should initially be completed annually starting at age 45, though may be done biannually after the age of 55.²⁰ Similarly for prostate cancer, the AUA recommends consideration for initiating testing for average-risk men between 45 and 50 years with PSA as first line screening test at a personalized screening interval. They also advise that DRE may be used alongside PSA to establish the risk of clinically significant prostate cancer.²¹ ACS recommends considering PSA testing at 50 years old, with timing for subsequent testing interval dependent on the value of the initial test.²² NCCN recommends considering PSA testing at 45 years old with consideration for baseline DRE, and further testing interval dependent on the initial baseline value.²³ For cervical cancer, ACOG, Society of Gynecologic Oncologists, and the American Society for Colposcopy and Cervical Pathology have all endorsed the USPSTF guidelines.²⁴ However, ACS recommends initiating screening with hrHPV alone every 5 years starting at 25 years through 65.²⁵ Thus, just as it is difficult for patients to know what to do with so much conflicting guidance, the same can be true for the individual cancer centres and providers trying to consolidate conflicting advice from multiple guidelines to provide advice for their patients. Healthcare providers, who ultimately have the discretion to recommend a specific screening strategy for their individual patient, may not rely on one resource but rather try to merge all the available evidence and guidelines with their clinical experience, and the context of patient and system factors in which they practice.

Additionally, we must consider the potential financial conflicts of interest when looking at cancer screening recommendations from individual cancer centers. Increased cancer screening, and all the downstream pathways that result from screening (extra imaging, biopsy, surgery, radiation, chemotherapy, etc.) are financially lucrative for cancer centers and provide an important stream of revenue for a healthcare system that is driven by volume, as is the case in the United States. Welch estimates that the United States’ expenditures relating to screening are in the territory of 40–80 billion USD per year.²⁶ Furthermore, with being a “top” cancer center, there can be a bias that “top quality” means more screening, newer modalities, more procedures. Indeed, several of these institutions also advertise “executive physicals” with an even more intense approach to cancer screening that is not be supported by evidence.²⁷ Thus, actions are needed at policy level to ensure consistent evidence-based screening guidelines across the centers, so as to prevent the public from being misled. It is important to note that more screening is always in the interest of the cancer centers, but not necessarily in the interest of the public.

Our findings are consistent with the work of Koh et al. who previously highlighted marked differences in prostate cancer screening between 607 US cancer center websites and USPSTF guidelines, specifically highlighting that some sites declare universal initiation of screening and the omission of an upper age limit to stop screening.²⁸ In addition, Clark et al. highlighted similar findings in lung cancer, noting that screening recommendations fail to highlight the benefits and harms in an equal manner.²⁹ Similarly, Patel et al. noted an earlier and more frequent use of mammography screening for BC than what was recommended by both the ACS and USPSTF, though the USPSTF recommendations have since changed compared to the recommendations available at the time, recommending early screening.³⁰

Policymakers could take several potential actions to ensure the consistency of screening recommendations across USPSTF and private cancer centers. For example, payment for cancer screening services could be limited strictly to USPSTF recommendations. The incentives to more screening could also be curtailed by including screening beyond the USPSTF recommendations as a poor quality/performance indicator. Additionally, the accuracy and consistency of recommendations could be factored into when deciding “top 10” lists of cancer centers.

Several limitations apply to the interpretation of our results. This is a cross-sectional study and thus, published information could have changed since our most recent review. Additionally, published cancer center recommendations may be reviewed and updated on different timelines compared to the USPSTF, which could explain some discrepancies. We also did not compare the cancer center screening recommendations against actual evidence as to whether these practices improve mortality, but rather used the current USPSTF recommendations as our “gold standard” comparison. We acknowledge that USPSTF guidelines may not be viewed as “gold standard” by all institutions, and are sometimes criticized, however these are the only national guidelines involving authors who are free from financial or professional conflicts. Our findings are based on an analysis of the Top Ten US Cancer Centers, which may limit generalizability to other institutions, including smaller cancer centers, community hospitals, or international settings with different healthcare policies and financial structures. These top-ranked centers may have unique institutional priorities, patient populations, and financial incentives that do not necessarily reflect broader national or international trends. We recognize that the published screening recommendations by cancer centers do not necessarily reflect the practice of individual physicians at these centers. However, it would be reasonable to believe that the cancer centers would not post on their websites what they wouldn't practice at their own institution. Finally, we don't intend to propose any screening recommendations as appropriate or inappropriate, which would depend on an individual's risk tolerance and preferences. We also acknowledge that conversations between healthcare providers and patients will contain more nuance than information published on a website. Our objective of this study remains simply to compare and contrast the publicly available recommendations of top cancer centers with the USPSTF guidelines and among themselves. Indeed, for screening, publicly available information on websites remain important because screening tests are sought out by people without symptoms and thus, not in contact with cancer centers yet.

In summary, our study found significant variability between the screening recommendations from the Top Ten US Cancer Centers and the evidence based USPSTF guidelines, especially for breast and prostate cancer, and to a lesser degree, cervical cancer. The discordance was usually in the direction of the cancer centers recommending more screening, often without discussion of harms. To address this issue, decisions need to be made at a policy-level to ensure consistency in the messages conveyed by governmental bodies such as the USPSTF, leading professional societies, and cancer centers.

Contributors

Kristin Wright and Bishal Gyawali were responsible for project conceptualization, methodology, and investigation. Kristin Wright was responsible for data curation and project administration. Kristin Wright and Brian Shkabari wrote the original draft which was edited by Bishal Gyawali. Bishal Gyawali provided mentorship and project supervision. Kristin Wright, Brian Shkabari and Bishal Gyawali verified data. All authors participated in review and editing of the manuscript, and have read and approved the final version of the manuscript.

Data sharing statement

All data used in this study were obtained from publicly accessible sources, including the USPSTF website and the official websites of the Top Ten US Cancer Centers. A full list of the cancer center webpages used for data extraction is provided in the Supplementary Appendix. The extracted dataset, including saved website content, can be made available upon reasonable request to the corresponding author. As this study does not involve patient data, ethical approval was not required.

Declaration of interests

Dr. Gyawali has received consulting fees from Vivio Health and holds equity options at OneCell Diagnostics, unrelated to the manuscript. Dr. Knopf has received advisory fees from and holds equity in Vivio Health, Cadex genomics, and One Cell Diagnosis. Other authors have no conflict of interest to report.

Appendix A. Supplementary data

Figure 6.

Color legend for Fig. 1, Fig. 2, Fig. 3, Fig. 4, Fig. 5.