Frequent loss to follow‐up after diagnosis of hepatitis C virus infection: A barrier towards the elimination of hepatitis C virus

Previous studies on hepatitis C cascade of care have been mainly focused on diagnosis and treatment rate, while less attention has been given to patients lost to follow‐up (LTFU) after diagnosis. Analyses of this latter issue on population level are missing.


| INTRODUC TI ON
Chronic hepatitis C virus (HCV) infection is a major cause of liverrelated morbidity and mortality around the world, with an estimated 71 million people infected. 1 The major route of transmission in industrialized countries is injection drug use. Therapy for chronic HCV has substantially improved with the introduction of interferon (IFN)-free direct-acting antivirals (DAAs) in 2014, enabling treatment with high cure rates, shorter duration and few side effects. 2 With new effective therapies, the World Health Organization (WHO) has launched a plan to eliminate HCV and HBV infections by the year 2030. 3 Elimination of HCV was defined as a reduction of new infections by 90% and a reduction of HCV-related mortality by 65% compared with the situation in 2015. To accomplish this goal, several factors must be in place for the HCV cascade of care, such as a high diagnosis rate of HCV infection, effective linkage to care, maintenance of patients in care, structure for scaling up treatment, and harm reduction to prevent reinfection and new infections. [3][4][5] Previous studies have shown that only a small proportion of persons infected with HCV reach the stage of antiviral treatment in the HCV cascade of care. [6][7][8] Much of the focus of HCV treatment has been the diagnosis of HCV, which can be challenging with different diagnosis rates between countries. 8 Another focus has been the low rate of treated patients among those diagnosed. 5,6 Less attention has been paid to the challenge of maintaining patients in HCV care after diagnosis or how to link the patients who were lost to follow-up (LTFU) back to HCV care.
Some studies of mainly small cohorts in different countries have shown that a high proportion of patients with diagnosed HCV infection are LTFU, likely because of the slow, asymptomatic and silent progression of the disease, in combination with the previous IFN-based treatment with substantial side effects and low chance of cure. 9,10 Retrieval of those LTFU back into the HCV cascade of care is therefore an important step on the road towards HCV elimination. Studies that address these issues are scarce, and analyses on a population level are missing.
In Sweden, the diagnosis rate of HCV infections is high compared with most other countries, but the treatment uptake has, until recently, nevertheless been low. 11,12 To implement the WHO elimination plan, there is a need to characterize the patients who were LTFU after diagnosis and to identify possibilities to retrieve them back to HCV care (eg, through other healthcare contacts). By using the unique Swedish personal identification numbers and the national registers with high coverage, a nationwide population-based study of people with HCV diagnosis and their healthcare contacts is possible. The aim of this nationwide study of people with HCV infection was to estimate the prevalence of those LTFU after diagnosis, to characterize them, and to analyze their non-primary healthcare contacts.

| Study population
This is a retrospective study of a national cohort of people with chronic HCV infection diagnosis in Sweden using prospectively The Swedish unique personal identification number (social security number) was used to link individuals between registers.
The NPR contains information on inpatient care in Sweden from 1964 (complete coverage from 1987), day surgery from 1997, and non-primary outpatient care (defined as visits to hospital-based medical specialists and will be referred to as specialist care) from Information on county of residence at diagnosis, highest attained education (≤9 years, 10-12 years, or > 12 years), marital status (married vs unmarried, divorced or widowed) at diagnosis, country of origin (Sweden, Europe or Non-European countries), and emigration/ immigration was added from the TPR and LISA registers.

Key points
• We showed in this nation-wide study that as high as two-thirds of patients with hepatitis C diagnosis were lost to follow-up from hepatitis C care after diagnosis in Sweden.
• The departments of surgery/orthopaedic or psychiatry/ dependency were the most common other non-primary healthcare contacts for patients lost to follow-up from hepatitis C care.
The patients with HCV infection were followed from the first

| HCV treatments
The PDR contains data of all prescribed drugs in Sweden from July

| Specialist care visits
In Sweden, most patients with diagnosed HCV infection will be referred to the Department of Infectious Diseases (ID), except those with decompensated liver disease that usually are referred to the Department of Gastroenterology and Hepatology (GH). Both ID and GH are specialized hospital settings (ie non-primary care) and the only departments/clinics that treated patients with HCV during the IFN-era, as well as being the only departments/clinics allowed to prescribe DAA treatment from 2014 to December 2017, hence limiting patients' access to HCV treatment to these departments/clinics. The departments of ID can be considered as secondary healthcare and not tertiary, since the patients are referred from diagnosing departments as primary care or other health care departments directly to these departments. As of January 2018, physicians at other departments were allowed to prescribe DAAs, but with the stipulation that this should be in collaboration with a physician at an ID or GH department. In this study, the ID and GH departments are therefore referred to as "HCV care departments." The inpatient and outpatient visits from 2001 to 2013 were analyzed using data from NPR. For patients who never attended HCV care during 2012-2013, other specialist departments/clinics were analyzed. These include all healthcare visits to specialized care in Sweden but do not include primary care. We analyzed the data in total and by each department. The non-HCV specialized care was divided into the following groups: Internal Medicine, Psychiatry/ Dependency, Surgery/Orthopaedics, Emergency department, ID/ GH contacts for other reasons than HCV infection (ie without ICD code for HCV), Dermatology/Venereology, Obstetrics/Gynaecology, Geriatrics and Other (Table S1).

| Statistical analyses
A univariate regression analysis was used to calculate the odds ratio for a follow-up visit at a HCV care department. Parameters

| The study population
In total 29 217 individuals with chronic HCV diagnosis in 2001-2011 were included for analyses (Table 1). Sixty-three percent were men, and the most common birth years were 1950-1969 (59%  When studying LTFU in different regions (by county), the estimates of LTFU varied from 49% to 65% (Table S2). These percentages were after implementing the general assumption that 15% already had been cured and did not need HCV care (applied equally in all Swedish regions).
The characteristics of patients attending HCV care (n = 9726) and those who did not attend HCV care (LTFU and cured; n = 19 491) are shown in Table 1. Since we could not identify the cured patients on an individual level, these patients were included in the group who did not attend HCV care when characterizing other specialist care contacts. sex. The visits to Obstetrics/Gynaecology decreased by age ( Figure S1).

| Predictors for LTFU and the effect of followup time
Male sex, younger age, lower education, psychiatric disease or dependency diagnosis, unmarried, HCV treatment and longer duration since HCV diagnosis were found to be significant risk factors for LTFU in the univariate analyses (Table 2). Living in a large city and country of origin (Sweden/Europe/Non-Europe) did not influence the risk for LTFU. All the significant factors in the univariate analyses remained significant in the multivariate analyses (Table 2).
Among the total study population, 58% (n = 16 889) did not see an ID or GH specialist for their HCV during the second year after diagnosis. Among the three quarters of patients (n = 21 924) diagnosed at ID/GH departments, 48% (n = 10 414) did not visit a HCV care department during the second year after diagnosis. This percentage was 89% (n = 6475) for patients diagnosed elsewhere.  Figure S2A). Among the patients LTFU, 70% (n = 13 662)

TA B L E 2 Univariate and multivariate analyses of risk factors for lost to follow-up after HCV diagnosis
attended an HCV care department during the first year after diagnosis, but this percentage dropped to 33% (n = 6497) during the second year after HCV diagnosis ( Figure S2B)  The IFN-based therapies were associated with substantial side effects and low cure rates, and some patients were treated several times without cure. These side effects and low cure rates have probably reduced the patients' interest in remaining in the HCV care continuum. Another important factor could be the long asymptomatic phase in combination with low awareness or knowledge of the risk for serious liver complications, leading to poor motivation to stay in the HCV care continuum. Besides missing the opportunity of cure from the infection, patients who are LTFU will not be monitored for liver disease. Therefore, they will miss the opportunity to be diagnosed for potential progressive liver disease and the subsequent prevention of liver-related death by surveillance for hepatocellular carcinoma or oesophageal varices with bleeding risks. We have previously shown the high risk of liver-related deaths in a nationwide study of patients with HCV in Sweden, emphasizing the importance of prevention of such mortality. 20 Liver biopsy, which is an invasive method with pain as the most common complication, has been previously used to assess the stage of liver fibrosis, and fear of such procedure has been identified as one of the patient-identified barriers to HCV care. 21 Nowadays, simplified tools for HCV care are available with the use of non-invasive methods such as transient elastography or algorithms based on blood markers to assess liver fibrosis, which together with the DAAs enable the monitoring and treatment for even hard-to-reach populations.

| D ISCUSS I ON
The HCV prevalence in Sweden is approximately 0.4%, with higher prevalence seen among men in the birth cohort born 1950-1969 and lower in older or younger age groups. However, transmission of HCV infection is still ongoing, and each year about 700 infections have been diagnosed in age groups of 15-29 years. 11 The drug abuse diagnoses were only found in 40% of the patients in this study, which is a low rate but this probably reflects that many past or current PWID are not in contact with health care for their drug abuse, and other arenas to reach the patients are needed. It has been estimated that up to 85% of HCV infections in Sweden are diagnosed and reported to the Communicable Disease Register. The diagnosis rate is a remarkably high proportion compared with most other countries. 12 This study shows that even in countries with a high rate of diagnosis, the treatment uptake can be low because of other obstacles in HCV continuum, such as the frequent LTFU.
This study showed that about two-thirds of patients did not attend HCV care as early as the second year after HCV diagnosis, emphasizing the need of early intervention after diagnosis to provide cure and reach the elimination goal. The predictors for patients LTFU in this study were young age, male sex, low education, presence of psychiatric or dependency diagnosis, unmarried, and longer duration since diagnosis of HCV. Young age may be associated with on-going drug use or lower motivation for follow-up because of the slow progression of the disease. Our findings emphasize the importance of providing easy and early access to HCV treatment for especially young patients or those with psychiatric or dependency diagnosis. early and easy access to HCV treatment and harm reduction for all patients, are necessary to reach the HCV elimination goal.

CO N FLI C T O F I NTE R E S T
SA has received honoraria for lectures/advisory boards from AbbVie, BMS, Gilead and MSD, and received research grants from AbbVie and Gilead. JS, JK and KB are employees of AbbVie and may hold AbbVie stocks or stock options. ASD has received honoraria for lectures/consultancies from AbbVie, BMS, Gilead and MSD.

AUTH O R S CO NTR I B UTI O N S
SA and AD conceived the study design, contributed to planning of the study and interpretation of data, and were the main authors of the manuscript. JS contributed to study design, planning of the study, data collection, interpretation of data, performed all the statistical analyses, and revised the manuscript critically. KB and JK contributed to study design, planning of the study, interpretation of data and revised the manuscript critically.