Peer Review Comments & Responses

Reviewer #1

Methodology:

The report includes all relevant studies and is complete.
I agree with the methodology and how the included studies and overall evidence have been interpreted and therefore believe the report correctly summarizes the available evidence.
It seems that pilot studies were differentiated from the remainder of the literature reviewed. Were these handled differently in terms of gathering evidence? The evidence retrieval process was the same for all studies, regardless of study design. However, an attempt was made to remove studies reporting on duplicate subjects, i.e., when the same subjects were reported upon in pilot studies and in final reports. Therefore, when two (or more) studies reported data on the same patient population for the same time period, only the study with the most recent or most complete data was kept. Pilot studies were not excluded from the review, although studies reporting estimated, projected, or self-reported data were excluded. To clarify, more details have been provided under “Literature Review.”
Although authors state that HCV RNA was not used as evidence of linkage to care, please consider giving more detail on how linkage was determined since this can be difficult to measure in many studies. More detail regarding ascertainment of linkage-to-care has been added under “Literature Review.” Abstraction of linkage-to-care data was harmonized to the extent possible, given the large heterogeneity between the studies regarding reporting linkage-to-care (e.g., studies not defining if linkage-to-care referred to establishment of an appointment or patient attendance at appointment, studies considering follow-up with primary care provider as linkage-to-care, studies not defining if treatment referred to initiation vs. completion of treatment, etc.)
Overall, scientific uncertainties are adequately addressed however consider adding more discussion about how the benefit of screening pregnant women may change if DAAs are approved for patients under 12. Understanding that this is speculation, introducing this as a possibility may be important to readers. The report has been updated to reflect that DAAs are now approved for children as young as 3 years of age. The report acknowledges that DAAs may become available for use during pregnancy in the future.

Recommendations:

Based on the evidence presented, I believe the CDC has come up with the right conclusions. Both recommendations are evidence based on a large body of literature.
Although stated in the body of the report, state prevalence of HCV should be more prominent. Providers may read these recommendations and wonder if they are living in a state with a >0.1% prevalence. Consider displaying this data in a box including the 3 states with <0.1% prevalence in pregnant women. A section entitled “Determining prevalence” has been added as well as language that emphasizes that health care providers should initiate universal screening in the absence of existing data. A new figure (Figure 2) has been added which displays state prevalence for adults and pregnant women.

Potential Impact and Implementation:

If implemented as written, I believe these recommendations would result in a reduction of the burden of hepatitis C in the United States.

Suggested Resources/Tools:

Recommendation of management of babies born to HCV positive mothers (when to screen and with what tests) Although outside the scope of this current recommendation document, CDC aims to separately assess the evidence informing hepatitis C testing recommendations for infants born to hepatitis C-infected mothers. This has been added under “Future Directions.”
Link to NAM HCV elimination reports The National Academies of Sciences, Engineering, and Medicine Reports (Eliminating the Public Health Problem of Hepatitis B and C in the United States: Phase One Report and A National Strategy for the Elimination of Hepatitis B and C: Phase Two Report) have been added and referenced under a new section titled “Diagnosis and Hepatitis C Elimination.”

Other Comments:

These recommendations are a large step in the right direction toward HCV elimination. Of course, more diagnoses will increase the need for linkage to care, medication access and capacity building to meet elimination goals.

Reviewer #2

Methodology:

The report includes all relevant studies and is complete.
The recommendation to screen in every pregnancy is not supported by the literature presented. There are no studies that examined the utility of screening in every pregnancy accounting for the fact that the vast majority of women will have tested negative with a first pregnancy and their subsequent risk of infection in future pregnancies is less than the population based prevalence, particularly in the setting of pregnancies (like most in the US) that occur close to one another in time. Additionally, the Tasillo study assumes 100% universal screening which is clearly not accurate given 1) our history with hepatitis B screening and 2) the fact that many women with HCV do not receive prenatal care. Although not supported by a cohort study, cost-effectiveness for screening during every pregnancy was supported by the Tasillo model. Furthermore, the Tasillo model might underestimate cost effectiveness because that model excluded the potential positive effects on infant outcomes and assumed 100% universal screening; a model assuming less than 100% universal screening may approximate risk-based screening and therefore overestimate cost effectiveness if persons with risk factors are screened at a disproportionately higher rate. Additional assumptions of the Tasillo model regarding linkage-to-care (25% linked-to-care at 6 months post-partum) and treatment have been added to the report. Hepatitis B screening occurs in 84-88% of pregnancies [Kolasa et al, PIDJ 2017], although screening is lower in subsequent pregnancies [Harris et al, Inf Dis Ob Gyn 2018]. Additional cost-effectiveness analyses of re-screening in subsequent pregnancies are forthcoming.
More details about the assumptions in the Tasillo study should be described. Additional assumptions of that model, including those related to linkage-to-care and treatment, have been added to the report.

Recommendations:

Based on the evidence presented, I believe CDC came to the right conclusions in some ways, but not in others.
I applaud the CDC for recommending screening of all adults. This will help to improve treatment prior to pregnancy and decrease the risk of fetal transmission as well as improve maternal health.
The recommendations for screening during every pregnancy are not well-supported. There is no alternative explored such as universal neonatal screening or having an any adult screen be adequate that does not need to be repeated during pregnancy. The cost of testing 4 million women per year, the vast majority of whom will be lower risk than the population prevalence due to prior negative testing is not small. For example, if a woman is tested just prior to pregnancy and then gets pregnant, should she be screened again? If a woman has 2 pregnancies in 1 year (or 2, or 3 years), should she be re-screened again? What duration of time should pass between screens if screening serially? Most cost effectiveness studies that I am aware of model a single screen. In my practice, if this recommendation rolls out it will not be rare to see a woman who has been screened for HCV 5 or more times over the course of 10 years. Since treatment isn’t currently recommended during pregnancy, the utility of serial screening seems questionable. Universal neonatal screening would likely not be feasible. For example, passively acquired maternal anti-HCV may be detected in infants until 18 months of age and obtaining a blood specimen from an infant can be challenging. Furthermore, this alternative would require screening of all infants, including those born to mothers who were not infected with HCV. The Tasillo model did include screening during each pregnancy and found this to be a cost-effective use of health care resources when simulating repeat pregnancy at a rate consistent with U.S. fertility patterns. The Tasillo model also likely under-estimated cost-effectiveness by not considering infant outcomes and assuming 25% of women were linked-to-care at 6 months post-partum. Although accepting results of testing that occurred prior to pregnancy may have the same health benefits as universal screening during pregnancy, challenges with retrieving past laboratory results likely make universal screening during pregnancy preferable, especially considering an obstetric panel of tests is routinely being performed. Because these recommendations are for screening during each pregnancy, there is no minimum duration of time between serial screens. However, providers may periodically reassess the hepatitis C prevalence in their patient population and may discontinue screening if it drops to less than 0.1% (e.g., as a result of prior identification and treatment of their HCV-infected patients).
The prevalence estimate seems like one that is going to be difficult to follow.This aspect of the recommendations may be challenging; however, unlike most other chronic communicable and non-communicable diseases, hepatitis C is curable. An effective hepatitis C screening and treatment program has the potential to significantly reduce disease prevalence in a population efficiently and quickly. Furthermore, once a disease becomes rare (the anticipated outcome of a successful hepatitis C screening and treatment program), effectiveness (including testing accuracy, and cost-effectiveness) is reduced. Therefore, this aspect of the recommendation is important from a public health standpoint. However, to help provide more guidance to providers, a section entitled “Determining Prevalence” has been added as well as language that emphasizes that health care providers should initiate universal screening in the absence of existing data. A new figure (Figure 2) has been added which displays state prevalence for adults and pregnant women.
Finally, the CDC should recommend timing of screening in pregnancy (if the recommendation persists) and be specific. Given that there are a number of recommendations based on cost-effectiveness studies, these recommendations should be based on those assumptions. Tasillo assumed screening during an early pregnancy visit in the economic model, which aligns with screening for other infectious diseases during pregnancy. However, screening later in pregnancy may be justified as this would identify women who acquire HCV infection later in pregnancy. Therefore, currently CDC is not making a recommendation on timing of screening during pregnancy although advantages and disadvantages of testing during different times are discussed under “Testing Considerations.”

Potential Impact and Implementation:

If implemented as written, I believe these recommendations would result in a reduction of the burden of hepatitis C in the United States.

Reviewer #3

Methodology:

The report includes all relevant studies and is complete.
I agree with the methodology and how the included studies and overall evidence have been interpreted and therefore believe the report correctly summarizes the available evidence.
The evidence findings of the report are clear.
The report clearly identifies and characterizes the scientific uncertainties and methodologic limitations and the potential implications of any uncertainties or limitations for the proposed recommendations.

Recommendations:

Since DAAs are not recommended for use during pregnancy, I am not entirely clear that repeat testing in each pregnancy is warranted. Furthermore, the recommendation about repeat testing during a single pregnancy for women at ongoing risk seems excessive without more compelling data. Unlike HIV, in which you can intervene during pregnancy to improve both maternal and neonatal outcomes, I am not sure the balance of evidence supports this level of testing during pregnancy. The cost-effectiveness of screening during each pregnancy is supported by the Tasillo model, which may have under-estimated cost-effectiveness by not considering infant outcomes and assuming only 25% of women were linked-to-care post-partum. HCV screening during each pregnancy is harmonized with other infectious diseases in which screening occurs during each pregnancy, which will likely lead to ease of implementation and obviates the need to track down prior laboratory results. Providers may periodically reassess the hepatitis C prevalence in their patient population, and may discontinue screening if it drops to less than 0.1% (e.g., as a result of prior identification and treatment of their HCV-infected patients).
The recommendations statement is clear.

Potential Impact and Implementation:

If implemented as written, I believe these recommendations would result in a reduction of the burden of hepatitis C in the United States.

Suggested Resources/Tools:

An app for clinicians including obstetricians.

Other Comments:

The document is carefully crafted and well written.

Reviewer #4

Introduction:

In second sentence, authors state: Percutaneous exposure is the most efficient mode of hepatitis C virus (HCV) transmission, and injection drug use is the primary risk factor for infection. While this is true, MSM are a high-risk group versus general population and this remains a lack of knowledge for providers, thus need to highlight risk in MSM as well. Language highlighting the risk for MSM has been added under “Persons at Risk for HCV Infection.”

New Recommendations:

How many providers know the prevalence in their community? Does this not provide a window for providers to assume their patient pop/community is low risk and not test? Real question is – is this provision really necessary? At this point, CDC will maintain the prevalence threshold although providers may continue to screen their patients universally for HCV infection even if their prevalence falls below 0.1%. This aspect of the recommendation may be challenging; however, unlike most other chronic communicable and non-communicable diseases, hepatitis C is curable. An effective hepatitis C screening and treatment program has the potential to significantly reduce disease prevalence in a population efficiently and quickly. Furthermore, once a disease becomes rare (the anticipated outcome of a successful hepatitis C screening and treatment program), effectiveness (including testing accuracy, and cost-effectiveness) is reduced. Therefore, this aspect of the recommendation is important from a public health standpoint. However, to help provide more guidance to providers, a section entitled “Determining Prevalence” has been added as well as language that emphasizes that health care providers should initiate universal screening in the absence of existing data. A new figure (Figure 2) has been added which displays state prevalence for adults and pregnant women.

Strategy to End the Hepatitis C Epidemic

3^rd paragraph, 2^nd sentence: remove the word “prior” Completed.
4^th paragraph: Should you include here the recent uspstf recommendation for universal testing, which would include preg women? The USPSTF recommendations are currently in a draft stage. If the USPSTF recommendations become finalized prior to publication of CDC recommendations, CDC will consider adding these.
5^th paragraph: The “2013-2016” is unclear, would reword. It is unclear what this time period related to. This has been clarified that this time period refers to the survey years.

Virus Description, Transmission, Clinical Features, and Natural History

4^th paragraph:
- Could be important to take 1 sentence to clarify that persons with HIV have lower spont clearance rates Completed.
- The following sentences should have references: Approximately 10%-15% of persons with hepatitis C will develop cirrhosis over 20-30 years. Those with cirrhosis experience a 1%-5% annual risk for hepatocellular carcinoma and a 3%-6% annual risk of hepatic decompensation, for which the risk of death in the following year is 15%-20%. Completed.
- Should last sentence include lymphomas? Completed.

Persons at Risk for HCV Infection

1^st paragraph, 4^th sentence: I think more recent data suggests this is more MSM and while HIV>non-HIV unclear how much greater risk is for heterosexuals with or without HIVThis sentence has been edited accordingly.
2^nd paragraph, 2^nd sentence: May be good to emphasize this increase by providing % increase and thus the increase in the number of children exposed A sentence has been added to address this.

Clinical Management and Treatment

1^st paragraph, 2^nd sentence: remove the word “generally” Completed.
1^st paragraph, 3^rd sentence: I don’t think this is true This sentence has been removed.
1^st paragraph, 4^th sentence should be edited as follows since non-nuc therapy is not part of current recommended regimens and the only regimen with non-nuc (dasabuvir) isn’t made anymore: The latest classes of antivirals for hepatitis C treatment include second and thirdnext-generation DAAs, categorized as either protease inhibitors, nucleostide analog polymerase inhibitors, non-nucleotide analogs, or nonstructural (NS5A) protein inhibitors.Completed.
1^st paragraph, 5^th sentence: Change ‘some’ to ‘many’ Completed.
1^st paragraph, last sentence: Add the following to end of sentence: , prior treatment experience, or presence of cirrhosis Completed.
2^nd paragraph, 1^st sentence: replace the word “lacking” with “preliminary and larger studies are required” Completed.
2^nd paragraph, 2^nd sentence: change to: However, testing women during pregnancy for HCV infection allows for testing of women who may only access health care during pregnancy and identification of infants who should receive testing This has been edited accordingly.
2^nd paragraph, 3^rd sentence: G/P is now approved for 12-17…would include this so as to be balanced Completed.
2^nd paragraph, 3^rd AND 4^th sentences: Led/sof is approved down to age 3 so 12-17 years should be changed to 3-17 years Completed.

Hepatitis C Testing Strategy

7^th sentence: some persons who test anti-HCV positive but HCV RNA negative could be in acute phase if they are high risk for recent exposure.This was not edited as RNA testing allows earlier identification compared to anti-HCV testing.

Recommendations

For children born to mothers with HCV infection should you specify appropriate age of testing and clarify appropriate test? Although outside the scope of this current recommendation document, CDC aims to separately assess the evidence informing hepatitis C testing recommendations for infants born to hepatitis C-infected mothers. This has been added under “Future Directions.”

Reviewer #5

Methodology:

The report includes all relevant studies and is complete.
I do not see data which supports the pregnancy recommendation for testing each pregnancy. I can see that the epi data would support testing a pregnant women who has not received the UNIVERSAL rec for adults but the “cost effectiveness considerations show that the first test would be similar cost to nonpregnant but tests in subsequent pregnancies would be less cost effective. The “linkage to care” considerations have no data related to pregnancy but if we extrapolate data from other pregnancy related morbidities and the ability to recapture those patients for gestational diabetes testing, less than a third of patients return for repeat testing and care. This doesn’t seem to support an every pregnancy rec either particularly when there is no medication that is deemed “safe” in pregnancy and there is no vaccine. Cost-effectiveness for screening during every pregnancy was supported by the Tasillo model. The Tasillo model might underestimate cost effectiveness because that model excluded the potential positive effects on infant outcomes. Furthermore, that model assumes only 25% of women are linked-to-care 6 months post-partum. The peri-partum period may represent a unique time in which women have access to medical care.
Overall the findings are clear. The one part that is confusing is the prevalence data – need to be much more concrete about how providers will know the prevalence in their community in order to determine using the test? To me the purpose of universal one time in adulthood is to capture people wherever they are regardless of risk factors given all of the pitfalls that exist with risk based screening. A new figure (Figure 2) has been added which displays state-level prevalence data among adults and pregnant women.
I am still a bit unclear about concerns in pregnancy with low prevalence particularly with unintended consequences that could occur with false positive testing in low prevalence situations? I may have missed that reference. The recommendations for universal screening do not apply when the setting prevalence is less than 0.1%. Persons who test anti-HCV positive should undergo reflex HCV RNA testing to determine if they are currently infected. The prevalence threshold applies to the pregnancy screening as well and the reviewer’s concerns are precisely why we believe the threshold is needed for all adults and pregnant women. CDC recommends reflex HCV RNA testing.

Recommendations:

Absolutely agree that universal screen of all adults including pregnant women if previously untested. I just don’t see data to support each pregnancy until there is safety data that we can do something during pregnancy when they are under our care or data to suggest that immediate knowledge to the baby is of use. Regarding the evidence, although not supported by a cohort study, cost-effectiveness for screening during every pregnancy was supported by the Tasillo model and additional studies are forthcoming. Regarding interventions, some interventions (e.g., avoidance of internal fetal monitoring, episiotomy, prolonged rupture of membranes and preference for amniocentesis over CVS) are recommended by the Society for Maternal-Fetal Medicine and inform pregnancy and delivery management. More text describing the justification for this recommendation has been added under “Clinical Management and Treatment.”
I would definitely clearly define practically how we determine the prevalence data. To help provide more guidance to providers, a section entitled “Determining Prevalence” has been added as well as language that emphasizes that health care providers should initiate universal screening in the absence of existing data. A new figure (Figure 2) has been added which displays state prevalence for adults and pregnant women.

Potential Impact and Implementation:

If implemented as written, I believe these recommendations would result in a reduction of the burden of hepatitis C in the United States.

Reviewer #6

Methodology:

For discussion of cost-effectiveness of population vs. birth cohort screening, please include paper by Coffin et al Clin Infect Dis 2012; 54:1259-71. They found that “compared to current guidelines, incremental cost per quality-adjusted life year gained (ICER) was $7900 for general population screening and $4200 for screening by birth year, which dominated general population screening.” This was added under “Cost-effectiveness Considerations.”
I agree with the methodology and how the included studies and overall evidence have been interpreted and therefore believe the report correctly summarizes the available evidence.
The evidence findings of the report are clear.
The report clearly identifies and characterizes the scientific uncertainties and methodologic limitations and the potential implications of any uncertainties or limitations for the proposed recommendations.

Recommendations:

Based on the evidence presented, I believe CDC came to the right conclusions.
The recommendations statement is clear.

Potential Impact and Implementation:

If implemented as written, I believe these recommendations would result in a reduction of the burden of hepatitis C in the United States.
Might be able to re-emphasize the importance of linkage to care and access to DAAs in discussion Completed.

Suggested Resources/Tools:

Downloadable slides with highlights, presentation at major medical conferences, press releases, coordination of screening efforts with public health departments. These tools will be considered as implementation and communication campaigns related to the new recommendations are developed.

Other Comments:

In the background section (page 5) “Clinical management and treatment”, the following comment is not true: “New drugs with different mechanisms of action and fewer negative side effects continue to become available.” There are no new antivirals in the US pipeline because the current DAAs are so effective and easily tolerated. I’d probably just delete this sentence. This has been edited accordingly.
In the same paragraph, last sentence, I’d suggest adding “regardless of HCV genotype and fibrosis level.” This has been edited accordingly.
The discussion about DAAs in pregnancy doesn’t account for presentation by Chappell et al at 2019 CROI, showing efficacy and safety in 10 pregnant women who received sofosbuvir/ledipasvir. This has been added.
In same paragraph (page 6), I’d change the statement “infected children can be monitored” to “should be monitored.” Chappell and colleagues showed very poor follow-up of these children born to infected mothers (Pediatrics 2018). Completed.
Methods: I like how the research questions were concisely and clearly laid out. Did the review process specifically include abstracts from scientific conferences? If so, this might be helpful for reader to understand comprehensive nature of review. Conference abstracts were not included, this has been edited accordingly.
It’s unclear whether ref 44 study was included in analysis or not. The paragraph describing this study seemed a bit of place. This has been deleted.
In the discussion about harms from screening for HCV in pregnant women, my colleagues in Ob/Gyn have cited the potential unintended consequence of more C-sections. I don’t know if this reason has been published, perhaps the Society for Maternal Fetal Medicine policy document has more detail (ref 42). This has been added.
Recommendations: The boxed recommendations are clear and thorough.
References: please doublecheck references for thoroughness and proper formatting (see ref 29, 39, 42). Completed.