How do (false) positively screened patients experience a screening programme for liver cirrhosis or fibrosis in Germany? A qualitative study

Abstract Objective This study aimed to explore psychosocial consequences of (false) positive liver screening results and to identify influencing factors for perceived strain within a multistage screening programme for liver cirrhosis and fibrosis in Germany. Methods Between June 2018 and May 2019, all positively screened patients were asked to participate in the study (n = 158). N = 11 telephone interviews and n = 4 follow‐up interviews were conducted. Semi‐structured telephone interviews were carried out. The analysis followed a structuring content analysis approach. Thereby, categories were first defined deductively. Second, the categories were revised inductively based on the data. Results The main themes found regarding the consequences of the screening were categorised in emotional reactions and behavioural reactions. Few respondents described negative emotional consequences related to screening. Those seem to be mostly driven by suboptimal patient–provider communication and might be worsened when transparent information transfer fails to happen. As a result, patients sought information and support in their social environment. All patients reported positive attitudes towards liver screening. Conclusion To reduce the potential occurrence of psychosocial consequences during the screening process, medical screening should be performed in the context of transparent information. Regular health communication on the side of health professionals and increasing patients' health literacy might contribute to avoiding negative emotions in line with screening. Patient or Public Contribution This study recognises the wide‐ranging patients' perspectives regarding the consequences of liver screening which should be taken into consideration when implementing a new screening programme to ensure a patient‐centred approach.

into consideration when implementing a new screening programme to ensure a patient-centred approach.

K E Y W O R D S
cirrhosis, fibrosis, liver screening, patient-reported outcomes, psychosocial consequences

| INTRODUCTION
Medical screening aims at identifying diseases in their preclinical phase to prevent severe progression. 1 Typically, screening is used to detect diseases before symptoms are present and thus, before patients seek medical advice for a specific problem. 2 Therefore, screening has the potential to move patients from a state of supposing themselves as healthy to the state of having a medical disorder. Reception of a positive test result represents a stress factor and can have severe psychosocial consequences. 3 This is especially an ethical problem, if screening results are false positive. [4][5][6][7][8] Besides the benefits of early detection of diseases, such as early treatment and potential prevention of progression, those negative effects should also be taken into consideration when evaluating new forms of screenings. 9,10 It is known from various studies on different screenings for cancer, that (false positive) abnormal findings in screening can lead to sleeping disorders, 4 increased anxiety, 4,11 psychological distress, 12,13 sadness, 12 restlessness, 12 fears 14 and considerations on future participation in screening. 15 However, those consequences do not occur consistently. It is important to differentiate between long-term and short-term effects 5,13 as well as between disease-specific and general outcomes. 8 In addition to false-positive results, intensive surveillance during the screening process itself can produce unfavourable side effects on psychological well-being and health-related quality of life due to the confrontation with a potential threat. 16 The majority of studies reporting on psychosocial effects of screening refer to different types of cancer screening programmes such as breast, 8 colorectal, 17 anal 18 and skin cancer. 19 Little is known about the impact of other medical screenings, for example, for advanced liver fibrosis or cirrhosis.
Cirrhosis is the common end stage of a chronic liver disease that often develops unnoticed over the years and thus is most often diagnosed in a late phase when complications occur. 20 In many cases, the transition between advanced liver fibrosis and cirrhosis is fluid. In this stage, causative treatment is less successful or impossible. 21 Furthermore, it constitutes a risk factor for the onset of liver cancer.
Even in highly developed health care systems, cirrhosis is diagnosed in an asymptomatic early stage only in about 25% of patients. 22 In January 2018, the SEAL programme (structural early-detection of asymptomatic liver cirrhosis or fibrosis) was implemented for 39 months in two German federal states (Rhineland-Palatinate and Saarland) aiming to investigate the feasibility, effectiveness and costeffectiveness of a general screening programme for liver fibrosis and cirrhosis in primary care. 21  For the SEAL programme, a cut-off value of 0.5 was chosen. In the case of a higher APRI in combination with at least one pathological transaminase, patients were considered positively screened patients with conspicuous liver values. Thus, a positive screening rate of 3.5%-4.0% was expected for the SEAL cohort with a false-positive screening rate at the first stage of 70%-80%. 23,24 Positively screened patients were referred to gastroenterological specialist examination for further clarification (second and third stage).
Within the SEAL programme, the present study was designed to explore potential psychosocial consequences, as well as behavioural changes for positively screened patients within the screening process. Since these patterns are complex and highly individual, a qualitative approach represents the best design to systematically explore all potential reactions and processes related to liver screening within the subjective reality of the concerned. Furthermore, with this study, we want to focus on the patient's perspective, which is often

| MATERIALS AND METHODS
The methods and results in this article are reported using the consolidated criteria for reporting qualitative research (COREQ) checklist. 25

| Participants and procedures
To identify positively screened patients, an interface to the electronic case report form (eCRF) was installed. The eCRF served as a management tool within the SEAL programme to collect all relevant patient data. When a new referral of a patient to a specialist was documented in the eCRF, the study team received an e-mail information.
Immediately afterwards, the patient was contacted by the first author via mail including a patient information and informed consent. He approached the interviewees with this information, including a statement that the study is independent of their medical treatment. The researchers were not in contact with the physicians or clinics of the patients and had no relationship before the study commencement. Incentives of €50 (in form of a transfer to a private account after the interview) were used to increase the response rate for this study since we received no responses within the first 4 months of recruiting. All patients, who were positively screened in the SEAL programme in the period between June 2018 and May 2019 were asked to take part in the study (n = 158). With a response rate of 7%, we could realise n = 11 telephone interviews.
Initially, a purposeful sampling strategy based on the criteria sex, age, comorbidity level and federal state was planned. However, since the recruiting phase took more than 1 year and the response rate was unexpected low, recruiting stopped after reaching interviews with n = 11 patients. The resulting convenience sample was based on the researchers' considerations that this sample size is (a) sufficient to reach data saturation and (b) feasible to analyse within the available resources. This rationale is supported by a work of Guest et al. 26 concluding that most themes emerge after 6-12 interviews. In some cases, a recall appointment (second stage) at the specialist was already scheduled at the time of the interview, hence a second talk after the consultation could be realised (n = 4). This option was offered to include further patient experiences even after correction or confirmation of a preliminary screening result to receive a more holistic impression. To keep the recruiting period in limits, we did not offer this for patients, who had no follow-up consultation scheduled.
To explore short-term psychosocial consequences homogeneously, we tried to realise the interviews as soon as possible after the initial mail approach. However, due to delays in communication and documentation in the eCRF, the time between physician visit and realisation of each interview varied.
For data collection, an interview guideline was developed based on the research questions. The guidelines covered four main topics: information about the screening, reactions to the results, external information retrieval and attitudes towards screening (see Supporting Information Material). This guideline was consented within a team of field experts (psychology, sociology, health services research and gastroenterology) and was pilot tested in the first interview. During the interviews, field notes were made to complement the data.
The telephone interviews consisted of two parts. First, an open narrative part, semi-structured by the guidelines, covered the main topics enabling the patients to speak openly. The second part was a short standardised query of sociodemographics, for example, age, education and comorbidities.

| Interview setting
All interviews were carried out by the first author (male), who has extensive experience in both qualitative and quantitative methods.

He holds a Master of Science in Sociology and Empirical Social
Sciences and was occupied as a research assistant at the University Hospital of Freiburg at the time of the interviews. At the beginning of each interview, the interviewer introduced himself, explained the goal of this study and repeated key information that was presented in the patient information sent out in advance. The interviewer emphasised that he has no medical profession and that this study is not related in any way to the medical treatment of the patients. All interviews were conducted by telephone while the respondents were at home. No presence of other cohabitants interfered with the interviews.
Transcripts and findings were not returned to participants for comment and or correction; however, the respondents were encouraged to contact the researcher after the interview in case of any upcoming thoughts or supplementary requests as a consequence of the interview.

| Data analysis
In sum, audio material of 4.5 h (approximately 19 min/call) was recorded. Audio data were transcribed verbatim by an external service provider and have been checked twice to ensure accuracy.
The analysis was conducted by the first author and followed the structuring content analysis approach by Kuckartz 27 using MAXQDA PLUS 2020 software. Thereby, categories were first defined deductively based on the research questions and assigned as part of the first coding procedure. Second, the categories were revised inductively based on the data, subcategories were formed, and a subsequent coding procedure was applied. When topics were addressed multiple times, they were also coded multiple times as text passages. The coding tree is available as Supporting Information Material.

| Analysis sample
In sum, nine women and two men were interviewed (see Table 1).
One person was accidentally included in the SEAL programme, though she did not fit the inclusion criteria for the main study (35 years or older). However, since this was not an exclusion criteria for the qualitative study, we decided to keep the interview data in the sense of a holistic approach. The mean age of the respondents was 65 years, however, since the mean is biased by the minimum extreme of 30 years, the median age (72 years) provides a better impression.
In comparison to the general SEAL population, our sample was, on average, 12 years older (median age of the whole SEAL population was 60 years) and less balanced regarding sex (SEAL population: 54.5% women). On average, three pre-existing conditions were mentioned with a minimum of one and a maximum of six. The majority of our sample reported to have low (54.5%) education (see Table 1) which is assumed to be associated with low health literacy. 28,29 The following four sections refer to the four research questions step by step.

| Perceived psychosocial consequences
Overall, the short-term negative consequences of screening were limited. With the exception of one patient (P 5), who reported that her liver values were good, each patient commented on their own emotions regarding excessively high liver values (see Table 2). Those were differentiated into negative and positive emotions, represented by two main categories.
In sum, five patients reported negative emotions about the screening results and five patients reported positive or at least neutral emotions in relation to the screening results. The negative emotions can be described as fear and sorrow as well as resignation (P 3, 8 and 11). In one case, the emotion was expressed as reduced will to live (P 4). However, the latter should not be attributed solely to screening since the person explained other circumstances that altogether led to this state (e.g., recent death of a family member). The positive emotions were characterised by comfort (P 1), happiness (P 2 and 10) and self-care (P 7). One patient used humour to answer the question (P 6).