We report on a nation-wide exercise utilising national diagnostic testing surveillance data and established clinical networks to re-engage individuals who previously tested positive for HCV and to offer treatment to those confirmed HCV RNA positive. This collaborative effort between UKHSA and the NHSE ODNs resulted in 7,442 (13% rising to 18% when we exclude previously treated, PCR negative and those who died) individuals who were previously not engaged in care being prescribed HCV treatment. These individuals are estimated to represent 1 in 10 of the total number of individuals treated in England since 2015. We also found that 2,990 (5%) individuals had died of whom 15% had HCC and/or ESLD recorded as a contributing factor on their death certificate. Overall the exercise was unable to directly re-engage 60% of individuals thought to be alive.
Our overall findings are similar to those published by Birmingham ODN which reported modest (11.3%) response rates to letters and low (25%) confirmed SVR numbers.(11) A similar exercise was conducted in Wales in 2019, with 23% of people contacted re-engaging with services, and 10% completing treatment in the first phase.(12) Similar studies have also been conducted in the Netherlands with 15% and 9% of those contacted treated(13,14) and in France where individuals with a previous HCV consultation but who had no record of cure were contacted, treating and curing 8%.(15) The Relink program used medical record review to identify eligible participants, re-engaged 33% of 11,163 participants in six countries, and treated 6%.(16,17) Similar to this exercise, the Trap Hep C programme in Iceland used cross-referenced surveillance and laboratory data and managed to re-engage all 24 participants achieving SVR12 for 83%.(18) However, the scope and scale of the problem is dramatically different in Iceland with an estimated viraemic population of 1,100 compared to 81,000 in England. (4) A clinical trial sought to evaluate the effectiveness of letters, phone calls, and other incentives to re-engage people previously diagnosed with HCV in the Canary Islands finding that phone calls were more effective than letters for re-engaging participants.(19) Participants that were not re-engaged were more likely to have a history of drug use, to have tested in the pre-DAA era, and were less likely to have a prior specialist evaluation.(20) Other patient tracing and re-engagement studies concerned with HIV care have had varying results from using a combination of methods to re-engage patients who are lost to follow-up,(21–23) including text messaging,(24,25) telephone calls,(26–28) email,(22) letters,(29) and physical tracing.(30–32)
Newly treated individuals were a mixture of individuals already known to the ODN, and individuals not known, who might have been unaware of their infection, aware of their infection but not engaged with healthcare services, and/or unaware of the emergence of new, better tolerated and more effective treatments. The number of individuals treated suggests that using national surveillance data as the basis for patient re-engagement exercises has some utility, but requires further interrogation of additional data sources to refine and validate the data, engagement from all stakeholders, with extensive follow up and local data checks by the ODNs. In some instances, despite positive re-engagement some individuals refused treatment. It is important that these individuals have continuing support and access to treatment should they change their mind.
Approximately 12% of individuals included in the exercise were found to be PCR negative with only a small proportion (4%) having evidence of treatment. This could reflect several mechanisms including spontaneous clearance of infection,(33) treatment outside the NHS (privately or outside England), or failure to record treatment. Given the mix of antibody and PCR tests in the laboratory surveillance dataset, the study likely included some individuals without an active infection. For example, a study conducted in GP practices in Southwest England found only 40% of participants recorded as antibody positive were confirmed to be vireamic.(34) Whilst of historic concern, the routine use of reflex PCR testing of antibody positive samples and point-of-care PCR testing in England(11,35) will likely make this issue obsolete for future similar exercises.
The exercise also revealed that 5.4% of individuals included in the exercise had died. A substantial proportion had HCV, ESLD and/or HCC as either a direct or contributory cause of death. The advent of DAAs has rendered HCV a curable disease in the vast majority of cases, so these deaths might have been avoided with earlier engagement in care. These findings illustrate the importance of the test and treat models, and ongoing work by NHSE to simplify the care pathway to ensure that people testing positive for HCV RNA have quick and easy access to treatment.
We found that 60% of those on the lists shared with ODNs could not be re-engaged. The majority of this group were males, 40 years or older and the largest proportion had been diagnosed between 2011–2015. Another study reported lower re-engagement for individuals diagnosed in the era preceding widespread use of DAAs.(20) In our study, varied levels of re-engagement likely represent implementation and individual-level challenges. Implementation challenges include varying ODN engagement with 11 of the 22 ODNs not reporting data, thus limiting our ability to fully evaluate the exercise. Secondly, there was significant heterogeneity in the implementation approaches used by ODNs. In the phase 1 evaluation, ODNs reported several obstacles including a lack of dedicated human resources and funding which may have contributed to this variability. Thirdly, due to varying data completeness, many individuals could not be found in any of the records ODNs cross-checked or could not be contacted due to a lack of up-to-date contact details. Studies suggest that better infrastructure could improve re-engagement exercises, including simple fixes like regular data sharing between health facilities.(36) Many printed letters, which were the main mode of contacting individuals in the exercise, were unanswered or not delivered, and letters have been shown to be less effective than other methods e.g., phone calls.(19) Finally, the exercise was designed to make successful re-engagement independent of GP involvement as GPs are often overburdened and have competing priorities.(37,38) However, the largest proportion of diagnoses in laboratory surveillance are made through primary care(39), and studies have shown higher treatment initiation and SVR rates for participants treated in primary care.(40) Closer integration of primary care could result in better outcomes as some studies in primary care have shown moderate success.(34)
Individual barriers could include internal, person-specific mechanisms such as anticipated stigma which can reduce the likelihood of engagement with healthcare services(41), with some of those affected by HCV being distrustful of institutions charged with their care.(42) Additionally, some of the population included in this study are highly transient and mobile (e.g., PWID who access drug treatment and rehabilitation services around the country, people living in prisons, people experiencing homelessness). Other studies have identified patient mobility as a major barrier to engagement with services.(43,44)
Re-engagement in care is challenging because prior disengagement is often tied to several factors including dissatisfaction with services, insufficient knowledge of HCV and treatment outcomes, complex needs, competing priorities and concerns about treatment side effects.(45) As such, better service delivery models that minimise barriers and maximise engagement opportunities, using an approach informed by behavioural science need to be considered. Services must adapt to cater to transient and underserved populations. For example, there has been a recent move towards patient-centred care with services offered through the most convenient route for service users.(46,47) For more complex cases, future exercises will require improved implementation and strong track and trace components to ameliorate their impact. Awareness campaigns are also necessary to educate the wider public about new well-tolerated, easy to take treatments with high cure rates and to enable people to objectively assess their own risk.(48)
There will also be a need to reconsider the design of the the re-engagement process which could be achieved through a qualitative evaluation and behavioural insights from ODNs and the population they serve. More effective engagement of ODNs may be achieved through monetary incentivisation which has been shown to be effective in engaging healthcare professionals.(49) In addition, ODNs might benefit from more effective networked data infrastructure that enables them to communicate and share data. This will require governance and data sharing agreements. Implementation could be improved with more quality control and refining of datasets before they are shared. ODNs would also require more support with data cross-checks, clinical team capacity building through training, and implementation toolkits to facilitate future exercises.
Strict criteria were used to minimise data errors in creating the lists provided to the ODNs and mitigate information governance risks. A further 42,426 individuals were excluded from the initial exercise due to lack of sufficient identifiers to permit robust linkage to surveillance datasets and 27,046 were excluded due to inconsistencies in their data. (Fig. 1) This highlights the importance of comprehensive data. Investment is needed to ensure that surveillance data are complete to allow for those requiring treatment to be identified, enable the monitoring of changes in service capacity and effectiveness, and improve our ability to track progress to elimination. Our results suggest that some of the individuals not included in the initial lists and not currently engaged in healthcare services might benefit from some additional attempt to re-engage them, which will impact HCV elimination goals.(2)
Reasons for the variation in ODN response are not well understood and merit further investigation. Qualitative in-depth interviews are planned to understand implementation choices made by ODNs, the burden and workload to ODNs, barriers and facilitators of the exercise, as this will provide insights into the patient care pathway and could optimise future re-engagement efforts. There is a pressing need to involve patients to understand the acceptability and reasons underlying choices to re-engage or not and to accept or refuse treatment. While ODNs indicated that the exercise appeared acceptable to patients and reported no adverse consequences,(9) qualitative research is planned to explore participants’ experiences. A separate exercise to re-engage children is also being conducted by the paediatrics team.
There are several key strengths of the re-engagement exercise. Firstly, the data used was diagnostic testing data for England which was made notifiable for HCV in 2010 and therefore should include all diagnostic tests for HCV from that time, as well as a majority of those reported prior to 2010. Secondly, we made use of the availability of several established healthcare databases to build the re-engagement lists which allowed for triangulation of data especially concerning HCV diagnoses, treatment and death, which was further enhanced by local checks undertaken by ODNs. Data linkage allowed for a more comprehensive range of information from multiple sources than would have been obtained had we relied solely on ODN reports.(50,51) As such, this should be routine practice in re-engagement exercises to ensure optimal acquisition of outcomes, to resolve uncertainties and to reduce bias.
However, the exercise also had some limitations. Firstly, there was varying engagement from ODNs. As such, our analyses were restricted by the amount and quality of data returned by ODNs. Secondly, information governance issues especially between laboratories and ODNs significantly hampered the exercise. Third, many individuals were included on the basis of a positive HCV antibody test, some of these may represent past rather than current HCV infection. Fourth, we cannot attribute all treatment initiations reported to the re-engagement exercise as this reason was not consitently recorded in the NHSE registry. Finally, it is important to acknowledge that the COVID pandemic caused significant disruption to the exercise.
In conclusion, this exercise was a substantial and extensive undertaking facilitated by access to key data resources and the participation of multiple organisations. The use of HCV surveillance data to re-engage individuals into care resulted in a sizeable number of people with known HCV infection accessing treatment. Further work is needed to investigate how those engaged differ from those who continue to have an unknown infection or treatment status. Repeat re-engagement exercises with improved implementation and alternative, complementary elimination strategies should be considered.