Emotional response to testing positive for human papillomavirus at cervical cancer screening: a mixed method systematic review with meta-analysis

ABSTRACT Tens-of-millions of women every year test positive for human papillomavirus (HPV) at routine cervical screening. We performed a mixed-methods systematic review using a results-based convergent design to provide the first comprehensive overview of emotional response to testing positive for HPV (HPV+). We mapped our findings using the cognitive behavioural framework. Six electronic databases were searched from inception to 09-Nov-2019 and 33 papers were included. Random-effects meta-analyses revealed that HPV+ women with abnormal or normal cytology displayed higher short-term anxiety than those with normal results (MD on State-Trait Anxiety Inventory = 7.6, 95% CI: 4.59–10.60 and MD = 6.33, CI: 1.31–11.35, respectively); there were no long-term differences. Psychological distress (general/sexual/test-specific) was higher in HPV+ women with abnormal cytology in the short-term and long-term (SMD = 0.68, CI: 0.32–1.03 and SMD = 0.42, CI: 0.05–0.80, respectively). Testing HPV+ was also related to disgust/shame, surprise and fear about cancer. Broadly, adverse response related to eight cognitive constructs (low control, confusion, cancer-related concerns, relationship concerns, sexual concerns, uncertainty, stigma, low trust) and six behavioural constructs (relationship problems, social impact, non-disclosure of results, idiosyncratic prevention, indirect clinical interaction, changes to sexual practice). Almost exclusive use of observational and qualitative designs limited inferences of causality and conclusions regarding clinical significance.

lesions compared with cytology-based testing alone (where cervical cells are microscopically examined for abnormalities) (Australian Government, 2017;US Preventive Services Task Force, 2018;von Karsa et al., 2015). Using HPV as the primary (first) test in cervical screening is considered to be the gold standard in many high-income countries and means that all women who attend screening receive an HPV-positive or negative result (Cuzick et al., 2006;Kitchener et al., 2009Kitchener et al., , 2014Rebolj et al., 2019). The Netherlands and Australia were first to fully implement HPV primary screening in 2017 (Australian Government, 2017;Aitken et al., 2019), and several high-income countries are in the planning, piloting, or early implementation stages (e.g., Sweden, Italy, UK, Norway, New Zealand) (National Screening Unit. New Zealand Government, 2017;Rebolj et al., 2019;Wentzensen et al., 2017). Other middle-and high-income countries, which have not yet switched to HPV primary screening, use HPV testing to triage borderline or lowgrade abnormal cytology (Arbyn et al., 2006). Globally tens-of-millions of women every year find out they are HPV-positive at their routine cervical screen.
Over the last few decades, the psychological impact of testing positive for HPV has attracted substantial research focus with many studies assessing emotional response, e.g., anxiety, concern about result, or worry about cancer. The rationale for research in this domain has usually been orientated towards attempts to mitigate unnecessary adverse psychological consequences (i.e., improve mental health outcomes) and to maximise screening re-attendance or help-seeking (i.e., improve behavioural outcomes). Given that cervical screening is usually a population-level intervention, assuming that HPV-diagnosis leads to even small percentages of women experiencing adverse effects, this translates to very large numbers experiencing negative psychological and/or behavioural sequelae. Hence efforts to monitor emotional response have been prioritised and commissioned through some national health bodies (Andreassen et al., 2019;Maissi et al., 2004;McBride et al., 2016). Despite research in this area however, to date, heterogeneity in local cervical screening protocols (e.g., screening tests used, order of tests) and study designs have meant that some major studies have produced mixed findings. For example, a large cross-sectional study found short-term anxiety and distress in women testing positive for HPV with abnormal cytology (Maissi et al., 2004(Maissi et al., , 2005. Qualitative research has also produced findings of anxiety, stigma, stress and concern about sexual relationships following positive HPV results (McCaffery et al., 2006;Waller, McCaffery, et al., 2007). However, a large randomised controlled trial which considered differences in anxiety and distress between women who were told their HPV-positive result vs. not told their result as part of routine screening practice found no overall differences (Kitchener et al., 2008). A qualitative study also reported indifference as a theme following HPV-positive results (O'Connor et al., 2014).
In addition to mixed findings, some psychological studies have adopted methodological designs using hypothetical scenarios (Brown et al., 2007;Kwan et al., 2010;Lee et al., 2007;Waller et al., 2009;Waller, McCaffery, et al., 2007). Since these studies ask participants to imagine their emotional response to testing positive for HPV, they lack ecological validity. Other studies have combined women with oncogenic and non-oncogenic HPV types, e.g., including women with genital warts (Graziottin & Serafini, 2009), or including women receiving treatment for precancerous cervical changes (O'Connor et al., 2015(O'Connor et al., , 2016. Again, this has meant that emotional response specific to testing positive for HPV at routine cervical screening has been difficult to isolate. Further, attempts to explain emotional response to HPV have been largely atheoretical to date. One study considered the role of illness representations and emotion in women with abnormal cervical screening results (without explicit HPV diagnosis), and found that emotion was explained by independent effects of a combination of demographic, cognitive and emotional representations (Hagger & Orbell, 2006). Leventhal's Common Sense Model (Leventhal et al., 2016) and Cognitive Behavioural Theory (Westbrook et al., 2011) have also been used by few studies to guide HPV-related interview or survey questions, reportedly proving useful frameworks (Maggino et al., 2007;Marlow et al., 2009). Speculatively drawing from theories and models of emotional adjustment, it is possible that appraisal and representations related to HPV diagnosis (e.g., sexually transmitted cause, lack of cure, perceived seriousness or control) (Folkman et al., 1986;Leventhal et al., 2016), concerns about cervical screening or treatment (Phillips et al., 2014), cultural/social norms and access to social support (Bandura, 1991), and coping or attachment style (Mikulincer & Shaver, 2008;Pietromonaco et al., 2013) may be important. Cognitive Behavioural Theory which underpins cognitive behavioural therapy (CBT), in particular, may act as a promising theoretical framework for provisionally mapping emotional responses and their related constructs. The CBT model encompasses interacting dynamics between emotions, cognitions and behaviours, and has been applied widely across health domains to identify overarching areas of importance for specific conditions (David et al., 2018). Whilst researchers working on psychological aspects of HPV are yet to establish a cogent theoretical framework, the CBT model may help organise relevant psychological responses and isolate areas for further concentrated theoretical developments. This is particularly relevant given that adverse emotional response to testing positive for HPV is likely linked to several other (potentially interacting) cognitive and behavioural outcomes (e.g., sexual relationships, health literacy, understanding of result). Research, however, is needed to establish which theoretical constructs are most relevant.
As it stands, there is a body of research on emotional response to HPV, but a lack of conclusive evidence which is useful for cervical screening programmes or informing theoretical advancement. Despite imminent roll-out of HPV primary screening in several countries and significant international interest, there has been no review or synthesis of the literature on emotional response. This mixed methods systematic review aimed to provide a comprehensive overview of the quantitative and qualitative literature, guided by the research questions: how do women emotionally respond to testing positive for HPV at cervical screening; and what influences emotional response to testing positive for HPV at cervical screening? Since emotions interact with, and are dependent upon, other biopsychosocial systems, the cognitive behavioural model (Westbrook et al., 2011) was also adopted to provide an overarching theoretical framework, which mapped the systematic review findings for emotional response into related themes of cognitions and behaviours. This helped formulate a preliminary working model of emotional response to HPV, in an otherwise predominantly atheoretical domain.

Method
This review is reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement (Moher et al., 2009) (See Supplementary File 1). The protocol was registered on PROSPERO on 15.08.2018 (reg: CRD42018105134).

Search strategy
Medline, Embase, PsycINFO, CINAHL, Global Health and Web of Science were searched to retrieve articles between 01. 01.1980 and 09.11.2019. The year coverage is representative of the earliest available database record until the date the last search was performed. The search concepts (HPV, cervical cancer, screening, psychological) were agreed a priori and informed by breaking down the research questions. The search strategy was developed for Medline, then validated and adapted for the other databases by an experienced librarian. Additional papers were identified by screening reference lists of included papers and searching OpenGrey (www.opengrey.eu). See Supplementary File 2 for the full list of search terms.

Design
We used a results-based convergent synthesis design, where the qualitative and quantitative evidence was analysed and presented separately then integrated by juxtaposing the findings in a matrix table (Hong et al., 2017;Pluye & Hong, 2014). For the purposes of this review, the integration synthesis was defined as refining, comparing and contrasting emotion-focused themes across all studies. Analysis of quantitative data estimated the relevance and representativeness of emotional responses by providing estimates of effect sizes and associations between testing HPV-positive and emotional outcomes; and analysis of qualitative data provided in-depth explanations for emotional response. See Figure 1 for an overview of this design.

Eligibility
The titles, abstracts and full-text papers generated from the searches met the following inclusion criteria:

Definition of emotion
Currently, there is no scientific consensus on an agreed definition of emotion. Popular theories, for example Plutchik's psycho-evolutionary theory of emotion (Plutchik, 2001), tend to be relatively consistent in how they describe primary emotions such as sadness, fear, happiness, disgust, surprise, anticipation, trust and anger. However, complex secondary and tertiary emotions, and their fusion with cognitions and physiological or behavioural cues, remain strongly debated across and within disciplines. Therefore, for the purposes of this review, we defined categories of emotion, and related cognitions and behaviours, based on a combination of the American Psychological Association (APA) published definitions (www.dictionary.apa.org/emotion), validated outcomes reported in papers, and the review team's interpretation in the coding and analysis stages.

Selection process
Extracted studies were included/excluded as part of a two-step screening process based on title/ abstract and full text. All titles and abstracts were screened by two reviewers (EM, OT or KW). Abstracts that passed the initial screen progressed to full-text review. Each full-text paper was independently assessed by two reviewers (EM, LR, OT) and discrepancies were resolved through independent full-text assessment from a third reviewer (JW), followed by discussion until consensus was reached. Agreement between reviewers prior to consensus was good (Kappa = 0.701). In some cases, authors of identified papers were contacted to request additional information where eligibility was not clear.

Data extraction
Data extraction was performed independently by two reviewers using customised Excel templates (EM, OT). Each reviewer's data extractions were compared and integrated to achieve the most comprehensive version. The information extracted across papers included: title, year published, study aims, sample size (total and by results group), population, study setting, participants (age, ethnicity, marital status and education), design, HPV and cytology results, outcome measures and analysis (where relevant), and main findings.

Data synthesis and meta-analysis
The data synthesis was conducted in three stages by two reviewers independently (EM, OT) with disagreements resolved through discussion or inclusion of a third reviewer until consensus was achieved (JW or ZR) (Thomas et al., 2004). Firstly, for quantitative studies, we assessed study designs, outcome measures and available data for inclusion in meta-analyses. We aimed to compare emotional responses in HPV-positive groups with a control group (e.g., HPV negative and/or normal cytology). Out of seventeen quantitative studies identified, six studies did not qualify for meta-analysis because their observational design did not include a comparison (control) group. A further three studies did not report data in a format suitable for inclusion in meta-analysis and corresponding authors were contacted in attempt to retrieve data; one author no longer had access to the data and two authors did not respond. Non-validated measures (e.g., singleitem questions) were also excluded from meta-analyses. From the available data, we were able to perform three meta-analyses for the outcome 'state-anxiety', and two meta-analyses representing psychological distress (by analysing outcome measures of general distress, sexual distress and testspecific distress together). We split the meta-analyses by time point (result notification ≤2 months [short-term] vs. >2 months [long-term]) and result group (HPV-positive with abnormal or normal cytology, vs. control). Statistical analyses were performed using Review Manager, version 5.2 (RevMan 5, 2012). Random effects models were chosen to account for heterogeneity in populations and design. Unstandardised mean differences with 95% confidence intervals were reported for anxiety as the included studies used the same outcome measure (STAI (Marteau & Bekker, 1992;Spielberger, 1983)). Standardised mean differences with 95% confidence intervals were reported for psychological distress as outcome measures differed between studies. Tests of homogeneity were conducted using the I 2 statistic (Borenstein et al., 2009). Low heterogeneity was depicted by I 2 values of <25%, moderate heterogeneity as 50% and high heterogeneity as >75% (Higgins et al., 2003). Tau-squared (τ 2 ) was reported to indicate estimates of between-study variance. We were unable to conduct meta-analyses for other emotional outcomes due to lack of data. See Supplementary File 3 for the raw data extracted for inclusion in meta-analyses.
Secondly, we synthesised all quantitative findings (including measures which could not be metaanalysed) by coding each measured outcome into themes of emotion, with related cognitive and behavioural themes also coded where relevant. Similarly for qualitative studies, the data were copied verbatim and thematic analysis was performed using descriptive and analytical coding to identify emotion themes, again with related cognitive and behavioural themes coded where relevant (Thomas & Harden, 2008).
Thirdly, to integrate the findings of the two syntheses (integrated synthesis stage), we refined the themes of emotion across the quantitative and qualitative studies. A conceptual matrix was then constructed by mapping the emotion themes by study, to allow for comparisons and contrasts. Narrative overviews of the quantitative and qualitative findings for each emotion-focussed theme are presented, with meta-analysis findings integrated.

Cognitive behavioural frameworkmapping interacting systems
Following the data synthesis stage, the cognitive behavioural model was adopted to provide an overarching and preliminary theoretical framework to map the findings into constructs of emotions, with related cognitions and behaviours (Westbrook et al., 2011). This helped address our second aim of understanding what influences emotional response to HPV. The cognitive behavioural model, which underpins cognitive behavioural therapy, was chosen because it has a strong evidence-base for explaining emotional response across psychology and health domains (Dobson, 2013;Hofmann et al., 2013). We used the model in its simplest form as a triad, to illustrate how emotions (feelings), cognitions (thoughts, beliefs, attitudes) and behaviours (actions) may interact to influence one another. In practice, this meant that alongside the primary thematic analysis phase, the qualitative verbatim data and quantitative outcome measures were also coded to represent constructs of cognitions and/or behaviours. These thematic constructs where then illustratively mapped onto the triad model of the cognitive behavioural framework. Two reviewers independently coded and analysed all data (EM, OT), with disagreements resolved through discussion or inclusion of a third reviewer until consensus was achieved (JW or ZR)

Quality assessment (risk of bias)
The Mixed Methods Appraisal Tool v2018 (MMAT) is a critical appraisal tool that has been specially developed for performing quality assessments in mixed method systematic reviews, and was used to assess the methodological quality of the included studies and potential for bias (Hong et al., 2018). The MMAT has independent sets of quality criteria to guide judgements for qualitative studies, randomised controlled studies, non-randomised studies, observational descriptive studies and mixed-methods studies. The quality score for each reviewed study was based on criteria specific to the study design, which included five methodological domains and was calculated as an overall percentage. Mixed-methods studies were assessed using the mixed-methods criteria as well as the separate quantitative and qualitative criteria; their quality score could not exceed the weakest component. We intended for the MMAT to be used for illustrative and descriptive purposes and did not weight findings based on quality score alone. Rather, each study was assessed independently on its merits, limitations and overall design in the cervical screening context by two reviewers (EM, LR, OT), with discrepancies discussed and resolved with a third reviewer (JW).

Rigour
Rigour was maintained by using a comprehensive search strategy along with documentation of eligibility decisions, which ensured descriptive validity (accuracy of data) (Sandelowski et al., 2006). Interpretive validity was achieved through use of at least two independent reviewers (EM, OT, LR) in the data extraction phase to create a comprehensive database and perform of quality assessments (Thomas & Harden, 2008). Following each stage of the data synthesis, two reviewers (EM, OT) plus a third reviewer (JW, ZR) discussed the thematic findings and resolved disagreements to help maintain theoretical validity (reliability of data interpretation) (Sandelowski et al., 2006). Pragmatic validity (efficacy and transferability of findings) was improved by inclusion of study characteristic tables providing the context around the studies, allowing readers to judge the usefulness of findings (Thomas & Harden, 2008).

Search results
The database searches yielded 15,792 papers, with 9,343 titles and abstracts screened after removal of duplicates. Ninety-three papers were fully screened and 33 papers, representing 32 studies, met (2) not outcomes of interest (e.g., HPV attitudes or knowledge without emotional outcome); (3) not empirical study; (4) no abstract; (5) HPV not in the context of cancer screening; (6) no clinical diagnosis of HPV (e.g., hypothetical scenario design); (7) only HPV vaccine related. 2 Phase 2 exclusion criteria described in the eligibility section used for full text articles.
the selection criteria. See Figure 2 for a Prisma Diagram providing an overview of the searches and selection process.

Study characteristics
Seventeen papers were quantitative studies (Alay et al., 2019;Andreassen et al., 2019;Ferenidou et al., 2012;Garces-Palacio et al., 2019;Hsu et al., 2018;Kitchener et al., 2008;Kwan et al., 2011;Maggino et al., 2007;Maissi et al., 2004Maissi et al., , 2005McBride et al., 2020;McCaffery et al., 2004;Nagele et al., 2019;Ngu et al., 2018;Rodriguez et al., 2019;Wang et al., 2010;Wang et al., 2011), 15 were qualitative (Barrera-Clavijo et al., 2015;Barreto et al., 2016;Bertram & Magnussen, 2008;Head et al., 2017;Kosenko et al., 2012;Lin et al., 2011;Linde et al., 2019;McCaffery & Irwig, 2005;McCaffery et al., 2006;McCurdy et al., 2011;O'Connor et al., 2014;Perrin et al., 2006;Tiro et al. (2019); Waller, McCaffery, et al., 2007;Wyndham-West et al., 2018) and one was mixed-methods (Daley et al., 2010). A total of 12,789 women aged between 18 and 65 participated in twenty studies (n=12,244 quantitative; n=545 qualitative), of whom 4,305 were reported as having tested positive for HPV (n=3,874 quantitative; n=431 qualitative). Seven studies were conducted in the UK, seven in the USA, six in China, two in Colombia and the remaining eleven in Australia, Austria, Brazil, Canada, Greece, Italy, Ireland, Mexico, Norway, Tanzania and Turkey. Twenty-one studies reported level of participant education: six used samples predominately educated to tertiary-level or above, and four primary-level or below. Fourteen studies reported a predominantly white ethnicity sample, and others predominantly African and Asian. Nearly all studies recruited women through clinical settings (e.g., hospitals, primary care), except two which used public advertisements and social media. Most studies ascertained diagnosis of HPV using clinical records; however, some relied on participant self-report. Time between participants receiving their HPV result and recruitment was not reported in the majority of studies (especially qualitative); but in those which did, the time from diagnosis ranged from shortly after receiving result (notification-2 months) to 2 years after result, with two outliers reporting 4.8 and 5 years. There were also variations in the combinations of HPV-positive and cytology result groups between studies: most used HPV-positive with abnormal cytology (any grade or mixed) and some used HPV with normal cytology, HPV with atypical squamous cells of undetermined significance, or HPV alone (no cytology test). Observational (cross-sectional, prospective longitudinal, or cohort) designs were used in most quantitative studies (thirteen out of seventeen). Four quantitative studies used a randomised controlled design (Garces-Palacio et al., 2019;Kitchener et al., 2008;Maggino et al., 2007;Ngu et al., 2018), but only one directly tested and reported differences in emotion between result groups (Kitchener et al., 2008). The same RCT study also included additional analyses on the observational findings from women in the study arm where participants were informed about their HPV results. All quantitative studies included at least one outcome with a core emotional component and most used widely-tested, validated scales; though some used single-item or non-validated scales and the mixed-methods study measured emotion descriptively. The most common outcomes measured were state-anxiety, sexual distress, testspecific distress, general distress, depression, fear and shame/disgust. Fourteen of the qualitative studies conducted interviews and one conducted focus groups (Barrera-Clavijo et al., 2015). All qualitative studies described at least one emotional theme, mainly related to anxiety, test-specific distress, sexual distress, surprise and confusion, fear, shame and disgust, sadness, relief and indifference.
Summaries presenting descriptive overviews of the studies and quality appraisal scores are presented in Tables 1 and 2.

Quality assessment
Overall, MMAT quality scores ranged from 40% to 100%. Qualitative studies scored highest for quality (median=100%, range 40-100%), followed by quantitative studies (median=60%, range 40-100%), and the mixed methods study (40%). The main reasons for quality deductions in the quantitative studies were non-complete reporting of data and not using appropriate measures; and in qualitative

Emotional response
We identified eight main themes of emotion which were measured or had emerged in women testing positive for HPV: anxiety; psychological distress (three types: sexual, test-specific and general); fear; surprise; shame and disgust; sadness; positive affect; and apathy. Each of these emotions are discussed separately with an overview of the synthesised evidence. See Table 3 for a brief definition these emotions. The main findings from the primary mixed methods study (Daley et al., 2010) were integrated with the relevant quantitative and qualitative components throughout. Tables 4 and 5 provide an overview of the main results for the quantitative and qualitative studies, respectively. Supplementary File 5 provides the integration matrix of the themes measured or emerged across all studies.
We were able to perform meta-analyses including seven out of eleven studies, comparing HPVpositive with abnormal cytology groups vs. control groups (normal or negative results) for both short-term anxiety (result notification ≤2 months) and long-term anxiety (>2 months). Results revealed higher short-term anxiety for women who were HPV-positive with abnormal cytology compared to the control groups across six studies (mean difference [MD] in STAI of 7.6, 95% CI: 4.59 -10.60, p < .001, τ 2 =11.11, I 2 =85%); however no differences were observed for long-term anxiety across four studies (MD = 0.03 95% CI: −1.45-1.51, p = 0.96, τ 2 = 0, I 2 = 0%). A small meta-analysis of three studies also compared HPV-positive with normal cytology groups vs. controls, which revealed higher short-term anxiety for HPV-positive with normal cytology (MD = 6.33, 95% CI: Table 3. Brief definition of each of the emotions identified as themes.

Anxiety
State anxiety describes an emotional state often characterised by apprehension, nervousness and/or uncertainty related to specific or future event(s) (Spielberger, 1983).

Distress
Psychological distress is a term to describe a collection of negative emotions or type of stress that results from being overwhelmed by demands or perceived threats. Distress can impact on everyday functioning related to general or specific events (APA, 2019b) Fear Fear is an intense basic emotion induced by perceived danger or threat(s) (APA, 2019c).

Disgust and shame
Disgust is characterised by strong aversion to something deemed unpleasant. Shame can stem from disgust and is characterised by a highly unpleasant feeling of humiliation or distress caused by the belief (or perception that others believe) that one has been dishonourable, immodest, or indecorous (APA, 2019f) Surprise Surprise is described as feelings of sudden unexpectedness. It results from violations of an expectation or detection of novelty in the environment (APA, 2019g), often followed by confusion.

Sadness
Sadness and depressive mood are usually temporary emotional states usually aroused by the loss of something that is highly valued (APA, 2019e). Clinical depression shares core characteristics with sadness but differs in that it is a serious longer-term mental illness which significantly impairs everyday functioning.

Positive affect
Positive affect is a broad and generic term for the internal feeling that occurs when a goal has been achieved, a source of a threat has been avoided, or one is satisfied with their current situation (APA, 2019d). Apathy Apathy is a lack of motivation, or the absence or suppression of emotion, interest, or concern and presents as a state of indifference (APA, 2019a).
To demonstrate the impact of HPV diagnosis on sexual function and mental health of Greek women.
Non-validated single item questions, except for sexual function which used a sexual dysfunction symptom checklist.
Majority (%) endorsed that they experienced anxiety (76.5%) after HPV diagnosis as well as fear regarding health in the future (82.4%). Nearly half of the women endorsed guilt (41.1%) and anger (43.1%). A minority endorsed distress, shame, reduction in self-esteem and stigmatisation (all < 22%).
Reduced sexual interest (33.3%) and frequency of sexual intercourse (43.1%) were also endorsed by some.

N/A N/A
Garces-Palacio et al. (2019) To assess the psychosocial impact of HPV testing, colposcopy and Pap-smear, as triage strategies after a Pap-smear with atypical Self-esteem; Anxiety; Psychosocial burden of HPV.
Rosenberg Scale (Rosenberg, 1989); STAI (Spielberger, 1983); Women testing positive for HPV with ASCUS had higher anxiety and psychosocial burden scores shortly after their result, Higher anxiety and psychosocial burden shortly after result but N/A squamous cells of undetermined significance (ASCUS) and evaluate the psychosocial impact based on the results of the strategies.
compared with HPV-negative women with ASCUS; however, there were no differences at 12months. Self-esteem scores did not differ shortly after result or at 12-months.
No effect for self-esteem.
Rodriguez et al.
To assess correlative factors that facilitate and inhibit transition to cervical cancer protective behaviour among women with HPV.
Stigma related to coping with HPV diagnosis and cervical cancer protective behaviour.
HIV Stigma Scale adapted (Berger et al., 2001 To assess the psychosocial impact of HPV testing as an adjunct to cytology in routine primary cervical screening. Anxiety (state and trait); General psychological distress; Sexual satisfaction.
STAI-40 (Spielberger, 1983); GHQ-28 (Golderberg & Williams, 1988); Sexual Rating Scale (Fedor-Freybergh, 1977) Women who knew they were HPV+ with mildly abnormal or normal cytology displayed no differences in anxiety or distress, when compared with those who did not know they were HPV+. Sexual satisfaction was lower in those who knew they were HPV+ with normal cytology, compared to those who did not know; but there were no differences for HPV + with abnormal cytology who knew vs. did not know. Women who knew they were HPV+ with normal cytology had higher state anxiety and distress, compared to those who knew they were HPVwith normal 1. RCT: no effect for anxiety or general distress; higher sexual distress for those with normal cytology only. 2. Cross-sectional revealed arm: higher state anxiety and general distress; lower sexual distress. S-STAI-6 (Marteau & Bekker, 1992); GHQ-12 (Golderberg & Williams, 1988); Higher state anxiety, distress and concern in HPV+ group compared to other test result groups.
Higher anxiety, general distress and test-specific distress.
Non-validated 2-item questionnaire (concern) predicted higher anxiety. No effect found for other demographic factors, awareness/ importance of HPV and perceived severity of cancer. Maissi et al. (2005) To describe the psychological impact on women of being tested for HPV when smear test results are borderline or mildly dyskaryotic at 6 month followup.
Anxiety (state); General psychological distress; Concern about result; Sexual health worries.
No differences for anxiety and distress. Higher concern about result and sexual distress To examine short-term anxiety and distress in women receiving different results following routine HPV primary testing at cervical screening. S-STAI-6 (Marteau & Bekker, 1992); GHQ-12 (Golderberg & Williams, 1988); Non-validated questions assessing concern, reassurance and worry.
Anxiety was significantly higher in women testing HPV-positive with either normal cytology or abnormal cytology, compared with the control group (normal cytology). Distress was slightly higher in women who tested HPV-positive with abnormal cytology, compared with the control group. There were also increased odds of very high anxiety (STAI score >49) in women who tested HPV-positive with normal or abnormal cytology compared to the control group. This pattern of results was only observed among women receiving their first HPV-positive result, not among women found to have persistent HPV at 12month follow-up. Odds of concern and worry were higher and reassurance lower, in HPVpositive groups compared to HPV-negative and normal cytology groups.
Higher anxiety shortly after HPV-positive with normal cytology (for first time) or abnormal cytology. General distress higher only for HPV-positive and abnormal cytology. Higher concern and worry and lower reassurance.
To examine the psychosocial impact of testing positive for high-risk HPV among women attending primary cervical screening.
Higher anxiety and test-specific distress in HPV+ with normal cytology, compared with HPVwith normal cytology. No differences in anxiety and test-specific distress for HPV+ with abnormal or unsatisfactory cytology, compared to HPV -Higher anxiety, test-specific distress and sexual distress.
N/A (Continued ) To examine the impact of different treatment strategiessurgical treatment or watchful waitingon sexual activity, psychosocial distress and fear of progression in women with HPV-associated premalignant genital lesions.
Fear of Progression; Sexual distress.
During an observational period of 12 months (baseline, 6, 12 months) there were no significant differences in fear of progression or sexual distress.
No effect over 12-months. N/A Ngu et al. (2018) To compare the effect of two educational interventions on the psychosocial wellbeing. Note: descriptive pre-intervention data and data from the leaflet (control) arm were extracted for this review.
Anxiety and Depression; Cervical cancer worry; Screening-related anxiety; HPV-related shame.
HADS (Zigmond & Snaith, 1983); CSQ (Wardle et al., 1995); Adapted Breast Cancer Worry Scale (Custers et al., 2014); Adapted STD-related shame questionnaire (Cunningham et al., 2002) Before randomisation to leaflet vs. counselling, 38.0% and 14.9% of women had clinically relevant anxiety and depression scores, respectively. Anxiety and cervical cancer worry were slightly lowered after receiving information in the form of a leaflet, but there were no differences in depression scores. Anxiety and cervical cancer worry decreased over time. There were no differences in HPV-related shame over time.

N/A N/A
Wang et al.
To describe the psychological impact of HPV.
HPV-Impact Profile (HIP) (Mast et al., 2009) Higher HPV-impact score in HPV+ with abnormal cytology compared to normal cytology.
Higher test-specific and sexual distress N/A Wang et al.
To assess the psychological burden of Chinese women with different HPV-related diseases.
HPV-Impact Profile (HIP) (Mast et al., 2009) Higher HPV-impact score in HPV+ with abnormal cytology compared to normal cytology. HIP domains 'sexual impact', 'selfimage' and 'control/life impact' had the highest scores. HPV+ with abnormal cytology showed sustained burden at 30 days, compared to HPVwith abnormal cytology which decreased.
Higher test-specific and sexual distress.
Psychosocial burden higher for women living in urban areas compared to rural. To evaluate the effect of communication and education strategies designed for women who participated in the comparative HPV testing and cervical cancer screening study, as an alternative technique to cervical cytology.
Anxiety, fear of cancer and fatalism in HPV-positive women. Also, blame towards partner. Face-to-face discussion with a health care professional reduced anxiety for many women. Barreto et al. (2016) To understand the feelings of women infected with HPV.
Fear, sadness and shame in HPV+ women. Bertram and Magnussen (2008) To describe the experience of women with abnormal Pap smears with a particular focus on their informational needs.
Initial anxiety at disclosure. Stigma associated with a sexually transmitted disease (STD) and a dearth of information available for male partners were problematic and influenced decisions about disclosure of human papillomavirus (HPV) infection to current or future partners. Daley et al. (2010) To assess the emotional impact and behavioural consequences following HPV diagnosis among women who had received abnormal Pap test results.

Head et al. (2017) To evaluate women's understanding of test results (Pap and HPV)
Confusion and anxiety in HPV+ women.
To determine the sources of uncertainty experienced by women living with HPV Seven sources of uncertainty: meaning of diagnosis; potential for disease progression; source of the infection; disclosure; sex and reproduction; and the HPV vaccine. Lin et al. (2011) To determine the psychological response of HPV infected women and their responses in terms of cognition, emotions and behaviour.
Linde et al. (2019) To understand causes of attendance and nonattendance to a follow-up cervical cancer screening among HPV-positive women.
Fear of cancer, confusion and relief that HPV was not cancer.

McCaffery and
Irwig (2005) To explore women's understanding of HPV, their information needs and experience of HPV infection using a method grounded in women's experience Anxiety and negative psychological response moderated by uncertainty about HPV, clinical communication and mode of delivery of result. Anxiety most associated with receiving the test result by letter and searching the internet for further information. McCaffery et al. (2006) To examine the social and psychological impact of HPV testing in the context of cervical cancer screening.
Stigma, anxiety, stress, concern about sexual relationships, and worry about disclosure. Psychological burden related to relationship status and history, social and cultural norms, and understanding of key features of HPV. McCurdy et al.
To examine Hispanic women's responses to learning they were HPV+, their decisions to disclose their HPV+ status, and their own and others' reactions to their disclosure.
All expressed surprise and fear; some expressed issues with disclosure. Higher concern expressed in single, unattached women under 28 years.
To explore emotional responses and predictors of negative reactions among women undergoing HPV tests in routine clinical practice.
Adverse emotional response (shame, embarrassment, stigma, regret, self-blame, anxiety, worry) linked to HPV infection rather than testing. Negative emotional response primarily influenced by concerns about abnormal cytology or diagnosis of CIN. Also, to a lesser extent, by HPV knowledge, awareness of HPV being sexually transmitted, awareness of HPV prevalence and HPV information needs. Perrin et al. (2006) To explore women's reactions to HPV diagnosis. Emotions related primarily to stigma, fear, selfblame, powerlessness and anger. Tiro et al. (2019) To explore patient perspectives after a positive HPV selfsampling result.
Main relevant emotional themes: intense affect after receiving positive results (e.g., fear of cancer and shock) and confusion about purpose and meaning of HPV testing. Also, relief after speaking to a healthcare professional and apathy (indifference). Waller, McCaffery, et al. (2007) To examine the way in which anxiety and concern transitioned over the course of the 12 months between two HPV tests; to explore the impact of a Adverse emotional impact (anxiety, shock, confusion, distress) reported initially for first test result. However, this did not generally last in the (Continued) 1.31-11.35, p = .01, τ 2 = 17.55, I 2 = 91%). It is worth noting that although the direction of effects were consistent across studies, high levels of statistical heterogeneity were identified in significant metaanalyses (I 2 >75%), therefore caution is warranted in the interpretation. See Figure 3(a)-(c) for the meta-analysis findings and papers included. Four studies which measured anxiety could not be meta-analysed due to study design (e.g., no suitable control group; (Ngu et al., 2018)) or lack of published data in the necessary format for extraction (Alay et al., 2019;Andreassen et al., 2019;Maggino et al., 2007). Consistent with the meta-analysis findings, two of these studies reported higher short-term anxiety in HPV-positive groups compared to controls (Alay et al., 2019;Maggino et al., 2007) but not long-term anxiety (Andreassen et al., 2019); and one study without a suitable control group found that anxiety decreased over time (Ngu et al., 2018).
Interestingly, a RCT which considered differences in anxiety between HPV-positive women who were told (revealed) vs. not told (concealed) their HPV status as part of an embedded trial in routine practice, found no differences between the groups (Kitchener et al., 2008). Predictors of anxiety in HPV-positive women were also explored in one study (Maissi et al., 2004): younger age, higher perceived risk of cervical cancer and not understanding the meaning of test results predicted higher anxiety within 4-weeks of results; but no predictive relationships were found for perceived importance of HPV and perceived severity of cervical cancer. year between the two test results. The emotional impact of a second positive HPV result 12-months later was greater for many women, sometimes causing them to disclose their result and seek support. Wyndham-West et al. (2018) To determine experiences surrounding HPV infections and pre-cancer.
Anxiety, shame, stigma, 'containment' of the infection (prevention), disclosure and social impact. Figure 3. (a) Forest plot comparing short-term anxiety (result notification ≤ 2 months) between those testing positive for HPV with abnormal cytology and control groups (HPV-negative and/or normal cytology groups). (b) Forest plot comparing short-term anxiety (result notification ≤ 2 months) between tose testing positive for HPV with normal cytology and control groups (HPV-negative and/ or normal cytology groups). (c) Forest plot comparing long-term anxiety (>2 months) between those testing positive for HPV with abnormal cytology and control groups (HPV-negative and/or normal cytology groups).

Qualitative (anxiety)
Ten qualitative studies reported anxiety as a theme following HPV-positive results (Barrera-Clavijo et al., 2015;Bertram & Magnussen, 2008;Daley et al., 2010;Head et al., 2017;Kosenko et al., 2012;McCaffery & Irwig, 2005;McCaffery et al., 2006;O'Connor et al., 2014;Waller, McCaffery, et al., 2007;Wyndham-West et al., 2018). Women who were anxious often had poor understanding of their results and/or HPV, expressed uncertainty about HPV, had often received their results by letter, and reported searching for further information on the internet (Head et al., 2017;Kosenko et al., 2012;McCaffery & Irwig, 2005;McCaffery et al., 2006;Waller, McCaffery, et al., 2007). Two studies found that women who had discussed their results face-to-face with a healthcare professional were less anxious (Barrera-Clavijo et al., 2015;McCaffery & Irwig, 2005). One study (Waller, McCaffery, et al., 2007) interviewed women after two HPV test results (12-months apart) and found that anxiety was a dominant theme shortly after a first or second HPV-positive result, but that it did not generally persist in the time between the two tests. A second HPV-positive test compared to a first one, however, was described as being more anxiety-inducing for some women.

Distress
Three forms of psychological distress were identified across studies: test-specific distress, sexual distress and general distress. Test-specific distress related to the psychological burden of HPV and screening test results. Sexual distress related mostly to impacts on sexual relationships, a partner, or concerns about transmission of HPV. General distress related to adverse impacts on everyday functioning (e.g., lack of sleep and concentration).

Quantitative (distress)
Sixteen quantitative studies included a measure of psychological distress: ten included test-specific distress (Ferenidou et al., 2012;Garces-Palacio et al., 2019;Kwan et al., 2011;Maissi et al., 2004Maissi et al., , 2005McBride et al., 2020;McCaffery et al., 2004;Ngu et al., 2018;Wang et al., 2010Wang et al., , 2011, eleven sexual distress (Alay et al., 2019;Ferenidou et al., 2012;Hsu et al., 2018;Kitchener et al., 2008;Kwan et al., 2011;Maggino et al., 2007;Maissi et al., 2005;McCaffery et al., 2004;Nagele et al., 2019;Wang et al., 2010Wang et al., , 2011 and six general distress (Andreassen et al., 2019;Hsu et al., 2018;Kitchener et al., 2008;Maissi et al., 2004Maissi et al., , 2005McBride et al., 2020). Test-specific distress was consistently higher (worse) for women testing HPV-positive with any cytology result compared to normal results up to 6-months post-result (Maissi et al., 2005), but not at 12-months post-result (Garces-Palacio et al., 2019). There were mixed quantitative findings for sexual distress and general distress. Sexual distress was found to be higher (worse) for women testing HPV-positive in five studies; however one low quality study showed no effect (Maggino et al., 2007), and another high quality found mixed findings depending on how they analysed their data (Kitchener et al., 2008). Another small study found lower sexual desire but no differences in overall sexual function between HPV-positive groups and the control (Alay et al., 2019); and a descriptive study reported that 33.3% and 43.1% of women endorsed reduced sexual interest and reduced frequency of sexual intercourse, respectively (Ferenidou et al., 2012). In terms of longer-term impact, sexual distress was found to persist at 6-months in two studies (Kwan et al., 2011;Maissi et al., 2005); one study examined the trajectory of adjustment to sexual distress over a 12-month period and found that adjustment occurred from one-to-6-months after HPV diagnosis (Hsu et al., 2018). Consistently, another study found no differences over a 12-month period (Nagele et al., 2019). General psychological distress (Golderberg & Williams, 1988) was found to be slightly higher (worse) in women testing HPV-positive with abnormal cytology 4-weeks after their result in two studies (Maissi et al., 2004;McBride et al., 2020). However, no differences were found 6-months later in a follow-up study (Maissi et al., 2005) or up to 12 or 24 months later in two other studies (Andreassen et al., 2019;Nagele et al., 2019). The Kitchener et al. (2008) trial again had mixed findings for general distress. Among women who were told their HPV result, being HPV-positive (vs. HPV negative) was associated with slightly higher general distress 2-weeks after the result. However, when women who had been told they were HPV-positive were compared with HPV-positive women who had not been told their HPV test result, no differences were found. Hsu et al. (2018) found that adjustment to general distress occurred between 1-and-6months after HPV diagnosis. We performed meta-analyses to combine the available data for test-specific distress, sexual distress and general distress to represent an overall measure of psychological distress in both the short-term (result notification ≤2 months) and long-term (>2 months). One study (Maissi et al., 2005) measured two forms of long-term distress (general and sexual); therefore, two meta-analyses were performed including each of these variables independently, to avoid bias through doublecounting in the total sample.

Qualitative (distress)
Themes indicative of test-specific distress emerged in thirteen qualitative studies (Barrera-Clavijo et al., 2015;Barreto et al., 2016;Bertram & Magnussen, 2008;Head et al., 2017;Kosenko et al., 2012;Lin et al., 2011;McCaffery & Irwig, 2005;McCaffery et al., 2006;McCurdy et al., 2011;O'Connor et al., 2014;Perrin et al., 2006;Waller, McCaffery, et al., 2007;Wyndham-West et al., 2018). Clear adverse impacts were reported, with many women describing concerns about HPV infection and/or the meaning of their test results. A small number of women reported that test-specific distress influenced their behaviours through triggering what they believed to be preventive action (often idiosyncratic, e.g., avoiding sharing soap/towels, exercising, or eating fruit) (Barreto et al., 2016;Wyndham-West et al., 2018). One study reported that test-specific distress primarily arose from concerns about abnormal cytology rather than HPV infection; however, it only included six women who were HPV-positive (O'Connor et al., 2014). The other studies reported that HPV infection had notably adverse impacts independent of abnormal cytology. Sexual distress was also a theme in nine qualitative studies (Barrera-Clavijo et al., 2015;Barreto et al., 2016;Bertram & Magnussen, 2008;Kosenko et al., 2012;Lin et al., 2011;McCaffery et al., 2006;McCurdy et al., 2011;Perrin et al., 2006;Waller, McCaffery, et al., 2007), with HPV-positive women describing a range of concerns about their sexual relationships, transmission of HPV and/or impact on their partner. Some women reported anger towards their partner and arguments due to suspected infidelity, or changing their sexual behaviours (e.g., avoiding sex) as a consequence of HPV.

Quantitative (fear)
Two studies descriptively reported that fear was a adverse reaction to HPV diagnosis, with 82.4% and 25% of women endorsing it descriptively (Ferenidou et al., 2012;Maggino et al., 2007). Similarly, the quantitative component of the mixed-methods study reported >75% endorsed fear; however, the authors categorised their definition of fear as endorsements of 'anxious' and 'worried' (Daley et al., 2010). Another study found that cervical cancer worry was higher in HPV-positive women shortly after result notification, but differences disappeared at 6-months (Kwan et al., 2011); and one study reported that worry about developing cervical cancer decreased over time (Ngu et al., 2018). During an observational period of 12 months (baseline, 6, 12 months) there were no significant differences in fear of disease progression (Nagele et al., 2019).

Qualitative (fear)
Fear emerged as a dominant theme in ten qualitative studies (Barrera-Clavijo et al., 2015;Barreto et al., 2016;Bertram & Magnussen, 2008;Lin et al., 2011;Linde et al., 2019;McCurdy et al., 2011;O'Connor et al., 2014;Perrin et al., 2006;Tiro et al., 2019;Waller, McCaffery, et al., 2007). Women mainly described fears related to the development of cervical cancer, their future health and potential infertility. Other women were afraid about the impact of their result or cancer on their family, partner and/or friends.

Quantitative (disgust and shame)
Six quantitative studies included measures of HPV-related shame or disgust (Daley et al., 2010;Ferenidou et al., 2012;Rodriguez et al., 2019;Ngu et al., 2018;Wang et al., 2010Wang et al., , 2011: two used the 'selfimage' domain within a distress measure (Mast et al., 2009); one adapted an STD-related shame scale (Cunningham et al., 2002); and two used non-validated measures. Shame and disgust were higher in women testing positive for HPV with abnormal cytology when compared to normal cytology (Wang et al., 2010, or HPV-negative with abnormal cytology  within 3-months of the result. Statements relating to shame and disgust were descriptively endorsed by the majority (>50%) in a descriptive study (Daley et al., 2010); and 'guilt', 'shame' and 'stigmatisation' were endorsed by 41.1%, 21.5% and 15.7% respectively in another study (Ferenidou et al., 2012). HPV-related shame did not change over time (up to 6-months post result) (Ngu et al., 2018), and one correlational study found that higher stigma was significantly associated with utilising fewer coping strategies and reporting less protective behaviour related to cervical cancer .

Qualitative (disgust and shame)
Shame and/or disgust also emerged as themes in eleven qualitative studies (Barrera-Clavijo et al., 2015;Barreto et al., 2016;Bertram & Magnussen, 2008;Lin et al., 2011;McCaffery & Irwig, 2005;McCaffery et al., 2006;McCurdy et al., 2011;O'Connor et al., 2014;Perrin et al., 2006;Waller, McCaffery, et al., 2007;Wyndham-West et al., 2018). These emotions mostly centred on concerns about disclosure of results to partner/family/friends, judgement from others and the belief that negative connotations (such as sexual promiscuity) were associated with HPV, sometimes leading to reports of stigma (McCaffery et al., 2006;O'Connor et al., 2014;Wyndham-West et al., 2018). Some women described feeling ashamed and reported variations of feeling 'unclean' or 'dirty'. Although shame and disgust appeared to be reported across different ethnic groups, these themes seemed more dominant in studies focusing on women from non-white ethnic backgrounds.

Surprise (and confusion)
Quantitative and qualitative (surprise) Despite surprise and/or confusion emerging as themes in ten qualitative studies (Barreto et al., 2016;Head et al., 2017;Kosenko et al., 2012;Lin et al., 2011;Linde et al., 2019;McCaffery & Irwig, 2005;Perrin et al., 2006;Tiro et al., 2019;Waller, McCaffery, et al., 2007;Wyndham-West et al., 2018), these responses were not measured using validated scales in any of the quantitative studies. One descriptive study reported that 70.1% of HPV-positive women endorsed that they felt 'shocked' (Daley et al., 2010). In qualitative studies, women often expressed surprise as the first emotion experienced after receiving their HPV-positive result. Many reported subsequent confusion about the meaning of HPV and how they had acquired it. Often surprise and confusion appeared to be linked with knowledge that HPV is sexually transmitted, raising questions about its source and concerns about potential infidelity (linking to sexual distress).

Quantitative (sadness)
One quantitative study descriptively reported that 14.9% of women who tested HPV-positive had clinically relevant depression scores; however, there was no control group to indicate population norms (Ngu et al., 2018). Another low quality study found that depressive/intrusive thoughts were slightly higher in women who tested HPV-positive compared to HPV-negative (time point not reported) (Maggino et al., 2007). A descriptive study reported that 51.7% of HPV-positive women endorsed that they felt 'depressed' (Daley et al., 2010).

Qualitative (sadness)
Only two out of eleven qualitative studies reported sadness or feelings of depression, and in both they were minor themes (Barreto et al., 2016;Waller, McCaffery, et al., 2007).

Quantitative and qualitative (positive affect)
In the quantitative studies, positive emotional responses, as indicated by improved outcomes following an HPV-positive result, were rarely observed. The only exception was one study where sexual satisfaction was higher in HPV-positive women (Kitchener et al., 2008). 'Relief' was also endorsed by 27.4%, 'encouraged' endorsed by 35.9%, and 'in control' endorsed by 68% of HPV-positive women in a descriptive study (Daley et al., 2010). Ten qualitative studies reported positive emotions such as relief, increased trust and acceptance, though they were minor themes (Barrera-Clavijo et al., 2015;Head et al., 2017;Kosenko et al., 2012;Lin et al., 2011;Linde et al., 2019;McCaffery & Irwig, 2005;Perrin et al., 2006;Tiro et al., 2019;Waller, McCaffery, et al., 2007;Wyndham-West et al., 2018). Women who reported positive emotional responses to their HPV results described receiving their test results in person by a healthcare professional, consulting with a healthcare professional after results, having a supportive partner and/or mobilising social support. Relief that the result was HPV and not cancer was a less common theme.

Quantitative (apathy)
One study descriptively measured apathy and found that 38% of women reported no reactive emotion to their HPV diagnosis (Maggino et al., 2007). Although the other quantitative studies did not directly measure indifference or apathy, the lack of observed differences in emotional outcomes between women receiving HPV-positive vs. negative results may be suggestive of apathetic or ambivalent responses, reported across quantitative papers under each individual emotion.

Qualitative (apathy)
Two qualitative studies reported indifference (O'Connor et al., 2014;Tiro et al., 2019), however this was either a minor theme or related more to the HPV testing procedures than response to testing HPV-positive.

Cognitive behavioural frameworkinteracting systems
The emotional response findings for quantitative and qualitative studies were additionally coded to identify related cognitions and behaviours, as a starting point to determine how these three factors interact. Within the eight broad emotion-focused themes, twelve cognitive constructs and ten behavioural constructs were identified (many of which are described in the results under each emotion).

Cognitions related to emotional response
Broadly, adverse emotional response to testing positive for HPV was linked to eight negative cognitions: low perceived control, confusion, stigma, relationship concerns, sexual concerns, cancer-related concerns, lack of trust in others and uncertainty about meaning of result or future health.
Conversely, neutral or positive emotional responses were linked with high perceived control, trust in others and acceptance.

Behaviours related to emotional response
Related to behaviours, six areas were linked to adverse emotional response: negative impact on relationships, negative social impact, non-disclosure of results, idiosyncratic prevention, indirect clinical interaction (e.g., results by letter) and changes in sexual behaviour. In brief, negative impact on relationships and negative social impact referred to themes such as reports of arguments with a partner or avoiding contact with others. Non-disclosure of results represented women who expressed that they deliberately concealed their result from others. Idiosyncratic prevention referred to reports of attempts to prevent the spread of HPV through engaging in activities that are not evidence-based, such as washing toilet seats. Indirect clinical interaction referred to receiving results by methods with no personal contact such as a mailed letter and/or not seeking advice from a healthcare professional. Changes in sexual behaviour described lower sexual activity, avoiding sex and/or using a condom.
Conversely, four behavioural themes were linked with positive or neutral emotional response: direct clinical interactions; social support; behaviour of others; and future screening attendance. In brief, women who reported speaking to a healthcare professional after their HPV-positive result (direct clinical interactions) or their partner/family/friends (social support) expressed feeling more reassured, less anxious, relieved and/or more accepting. Helpful behaviours of others related to partners/friends/family sourcing information on HPV or encouraging help-seeking behaviours. Attendance at a screening appointment after receiving an HPV-positive result (future screening attendance) was described by some women as providing reassurance.
According to the cognitive behavioural model, these three constructs of emotions, cognitions and behaviours are likely to directly influence and/or interact with one another. This formulates a working model of what may influence emotional response to testing positive for HPV. See Figure 5 for an overview of emotions, cognitions and behaviours mapped on to the cognitive behavioural framework.

Discussion
This systematic review provides a comprehensive overview of emotional response to HPV diagnosis at cervical cancer screening, as well as a provisional model for understanding how emotions may interact with cognitions and behaviours using the cognitive behavioural framework. Testing positive for HPV at cervical screening appears to be most strongly associated with short-term anxiety, short and long-term psychological distress, and related to feelings of disgust and shame, surprise and fear about cancer. There was little evidence of sadness or depression and a minority of women reported apathy or relief that they had been diagnosed with HPV rather than cancer.
Anxiety was one of the most common adverse responses reported shortly after women had received their HPV-positive result across all studies. Our meta-analyses revealed higher short-term state anxiety in women testing positive for HPV with abnormal cytology or normal cytology when compared with normal screening results (mean difference on STAI (Spielberger, 1983) of 7.6 and 6.33, respectively); though high statistical heterogeneity was observed, potentially due to differences in screening contexts and magnitudes of effect sizes (I 2 >75%). These findings are consistent with  Figure 5. Emotional response to testing positive for HPV from all studies (quantitative, qualitative, mixed-methods) mapped on to a cognitive behavioural framework another systematic review which found elevated anxiety in women with abnormal cytology who were attending for colposcopy (a more advanced stage in the screening process) (O'Connor et al., 2016). Interestingly, when comparing our results to this review, anxiety scores observed in colposcopy patients appeared to be descriptively similar to women testing positive for HPV with abnormal cytology (mean STAI score range: 34.0 -49.0 pre-colposcopy vs. 39.6 -46.0 after test result). These similarities suggest that anxiety associated with an HPV-positive screening result may be comparable to the anxiety experienced at follow-up investigative procedures (colposcopy); or may persist from the time of result to colposcopy.

Cognitions
Reassuringly, however, the results from our meta-analysis revealed that anxiety did not appear to persist in the long-term (> 2 months after notification), when comparing HPV-positive with abnormal cytology vs. normal/negative result groups. Also, overall, the mean anxiety scores observed across studies did not generally exceed thresholds for clinical significance. The anxiety scores associated with a HPV-positive result tended to be higher than expected in the general population but lower than the cut-off for clinically important anxiety. Although, it is worth noting that all quantitative studies assessed anxiety across the whole study sample without conducting subgroup analyses. From a clinical perspective, it is highly unlikely that acute adverse emotional response to HPV would be expected or detectable at the population level. It is more likely that certain groups of women would be at higher risk of clinically important anxiety (e.g., low socioeconomic status, ethnic minority groups, low health literacy) who should additionally be studied or analysed separately. Anxiety was a dominant theme in the qualitative literature which, due to the likelihood of self-selection bias in qualitative studies, supports the notion that certain groups of women may be prone to very high anxiety.
HPV positivity was also related to psychological distress in both the short-term and long-term. Our meta-analyses (which combined sexual, test-specific and general distress) revealed higher distress in women testing HPV-positive with abnormal cytology when compared with normal/negative results, at both result notification to 2-months and 2-months onwards. Long-term distress (> 2-months), however, seemed to be specific to sexual and test-specific distress, as the studies which measured general distress at this time point found no differences.
Experiencing distress related to sexual relationships, infidelity and potential transmission of the virus (sexual distress) is consistent with the broader literature on emotional response to other STIs and HPV in non-screening contexts (e.g., genital warts, other cancers) (Dodd et al., 2016;Graziottin & Serafini, 2009). In this review, sexual distress appeared mostly, but not exclusively, limited to women in relationships and/or with current sexual partners in the qualitative literature, which may help explain some heterogeneity in findings. For some women, it was also reported as associated with relationship problems (e.g., arguing over suspected infidelity) and changes in sexual practice (e.g., avoiding sex).
Distress related to the meaning of screening test results (test-specific distress) was very common in the qualitative literature and was often described as the successor to surprise and confusion. It was mostly linked to low HPV awareness, not understanding result meaning, confusion about the aetiology of HPV and concerns about future health. As HPV cannot be cured and there are no clear (practical) prevention methods available (except vaccination prior to exposure), some women reported feeling that they were not in control of their health. Low perceived control appeared related to higher test-specific distress. A small number of women also reported engaging in idiosyncratic prevention methods to help treat or 'contain' HPV, such as washing toilet seats or increasing physical activity. As a psychological formulation, these forms of prevention could be interpreted as behavioural attempts to gain control and reduce distress (Westbrook et al., 2011). High levels of distress about result also appeared to be closely related to fears about developing cancer which, together, intensified overall adverse emotional response.
Shame and disgust emerged as themes in the qualitative studies and a small number of women also reported feeling that there was stigma attached to HPV, which is consistent with broader STI research (Bickford et al., 2007;Jeynes et al., 2009;Nack, 2000). In line with sexual distress and test-specific distress, shame and disgust seemed to be associated with maladaptive behaviours. Some women reported reluctance to disclose their HPV result to others and/or to seek social support from their partner, family, or peers because of feeling ashamed. To further assess the relevance of shame and disgust in the cervical screening context, future quantitative research should incorporate validated measures which include relevant behavioural impacts.
Relatively few studies measured sadness, depression, or generalised distress. In those studies which did, there was little evidence of adverse (clinically important) effects associated with any HPV-positive result. A small number of qualitative studies reported positive or neutral emotional responses, such as relief that a test result was HPV and not cancer or indifference. However, these were not common and/or dominant responses.
Across all studies (quantitative and qualitative), adverse emotional response was mainly related to not understanding the meaning of the result, being in a relationship or having a current sexual partner, non-white ethnicity, receiving test result by letter, not discussing the result with a healthcare professional, little social support and lower levels of education. Adverse emotional response was observed across all studies but appeared most prominent in the qualitative literature. Although fear and surprise/confusion were common themes in the qualitative studies, they were rarely measured in the quantitative studies, highlighting a gap in quantitative research which warrants further exploration. Overall, our findings suggest that receiving an HPV-positive result at cervical screening can cause significant disturbance for some women, however, likely the minority of the population and/or certain groups.

Methodological considerations
Importantly, this systematic review raises some relevant methodological considerations. Nearly all studies adopted cross-sectional, descriptive and/or qualitative designs, prohibiting inferences of causality between testing positive for HPV and emotional response. The persistence of HPV infection (and the development of abnormal cells) are closely intertwined with immunological response; and there is a body of literature which suggests that psychological or social stressors can impair immune response (Fang et al., 2008;Marsland et al., 2017;Steptoe et al., 2007). Therefore, it cannot be ruled out that HPV activation and/or persistence are functions (or sub-functions) of psychological stress (i.e., adverse emotion). Interestingly, the one large RCT study in this review which compared anxiety and general distress between women testing HPV-positive who were told (revealed) vs. not told (concealed) about their HPV status (Kitchener et al., 2008), found similarly elevated anxiety scores (no differences). This suggests that elevated levels of anxiety and distress may be present prior to learning HPV-positive screening results, which supports the notion that psychological stress could play a role in HPV activation/persistence. Other research suggesting that anxiety associated with HPV is usually temporary and normalises at 6-month follow-up may provide evidence against this mechanism; although it is worth noting that 41% of HPV cases clear within 6 months (Bulkmans et al., 2007), meaning effects may be confounded. Further research is needed to test the validity of such psychobiological mechanisms and/or other potential causative pathways.
Very few studies also analysed and/or interpreted their data in terms of clinical significance, meaning it was not possible to distinguish between normal and clinically relevant emotional responses for most outcomes. Negative response to adverse information is usually a temporary process constituting a normal part of human consciousness. Therefore, studies in this review which drew implicative conclusions based on between-group differences without further interpretation provided little insight distinguishable from healthy response. To progress this field of psychological research, future studies should be designed and appropriately powered to test for clinical significance rather than between-group differences alone.
Most participants were educated to secondary level or above (where it was reported) and there were relatively few studies from low-and-middle-income countries. The highest quality studies consisted of well-educated (tertiary level) white patients living in high-income-countries which used organised screening programmes. Consequently, the main findings of this review are weighted towards relatively homogenous samples and may not be directly translatable to other settings or lower-income-countries. The qualitative studies which were conducted in low-and-middle-incomecountries (Brazil, Colombia, Taiwan, Tanzania) reported stronger adverse emotional impacts (Barrera-Clavijo et al., 2015;Barreto et al., 2016;Lin et al., 2011;Linde et al., 2019). Therefore, the findings reported in this review may be conservative compared to other health systems or cultural contexts.
Finally, HPV-positive results in the reviewed studies were usually accompanied by abnormal cytology. This meant that we were unable to determine the relative impact of HPV vs. abnormal cytology for many of the emotions described. However, there were some emotions which seemed inherently related to HPV, such as sexual distress and test-specific distress. Receiving both HPV and cytology results is, nevertheless, reflective of routine screening practice, meaning that the findings of this review should provide valuable and pragmatic insights into the patient experience at screening.

Limitations
Our timely systematic review benefits from the adoption of a relatively novel and rigorous mixed methods review design. Like most reviews, we have a number of limitations worth considering when interpreting the results. Firstly, although we used a comprehensive search strategy to identify papers across six major databases, our grey literature search was limited to OpenGrey and we did not contact authors or Listservs to identify additional literature. Also, given that there is no clear agreed or distinct theoretical definition for many emotions, emotion categorisations were often based on judgements and interpretations by the review team, especially where data was measured using non-validated scales or qualitative data. The meta-analyses were also performed using small numbers of studies (range: 3 -6) which can be unreliable and subject to bias, and prohibited moderator analyses. Therefore, relevant mechanisms could not be explored and caution is warranted in meta-analysis interpretations. Lastly, whilst we used the cognitive behavioural framework to map our findings, there are several other potentially more relevant theoretical models which could be used to structure emotional reactions to HPV; e.g., Williams' Affect and Health Behavioural Framework (Williams & Evans, 2014) or Leventhal's Common Sense Model of Self-Regulation (Leventhal et al., 2016). Using alternative theoretical frameworks may have led to different formulations but we are confident that our overall conclusions are valid.

Implications for policy and practice
As HPV primary screening is being implemented around the world, our findings provide rich insight for policymakers and clinicians into women's experience of receiving HPV-positive results. In attempts to mitigate adverse response, common themes highlighted in this review (e.g., related to confusion around cancer risk or sexual transmission) could be targeted through tailored information in screening result letters or accompanying leaflets. Clinicians working in primary care and cervical screening in areas where HPV-testing is being implemented could also use this information to preempt or address women's questions and concerns, especially in low-and-middle-income countries where adverse emotional response may be greater. Public health or third sector organisations running campaigns on cervical cancer screening could frame their communications to target some of the key areas, e.g., to tackle stigma associated with sexually transmitted aspects. Clinical signposting and pathways could also be embedded within cancer screening programmes to provide support for some of the sub-groups highlighted, who may be at higher risk of clinically important adverse responses (e.g., women from ethnic minority backgrounds, or those with low health literacy or without access to social support).

Conclusion
Short-term anxiety, distress about test results, distress about sexual relationships, feelings of disgust and shame, surprise, and fear about cancer appear to be the most common emotional responses to testing positive for HPV. Almost exclusive use of observational and qualitative designs, however, limits conclusions regarding clinical significance and prohibits some important causal inferences. We hope this comprehensive review, paired with our provisional framework of relevant emotion, cognitive and behavioural factors, will act as a springboard for the development of a cogent theoretical literature on this topic.

Author contributions
EM, ZR, OT and JW conceived the study. OT, KW and ZR conducted the searches. EM, OT, LR and JW selected eligible studies. RMM, LM and NK provided intellectual input on the review design. EM, OT, LR and JW assisted with data extraction and quality assessments. EM, OT and JW conducted the syntheses and analyses. EM drafted the paper. All authors contributed to the final version of the manuscript. The corresponding author attests that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted.

Disclosure statement
No potential conflict of interest was reported by the author(s).