Quality of life, anxiety and depression patient‐reported outcome measures in testicular cancer: A systematic review

Several patient‐reported outcome measures (PROMs) are available for the assessment of quality of life (QoL), anxiety and depression for testicular cancer (TCa); however, these PROMs have uncertain validation of their psychometric properties for TCa‐only cohorts. This systematic review aims to critically analyse and evaluate the psychometric properties of these QoL, anxiety and depression PROMs.


| BACKGROUND
Testicular cancer (TCa) is the most common solid malignancy of men between the ages of 14 and 44 1 and represents 5% of urological cancers and 1%-2% of all male neoplasms. The global incidence of TCa is increasing, with 3-10 new cases occurring per 100,000 males/per year in European countries. 2,3 Earlier detection (e.g., by means of heightened public awareness relating to testicular self-examination), young age at presentation, high healing rate of testicular neoplasms and surgical technology have resulted in a high median five-year survival rate for TCa of ∼95% and achievability of complete remission in metastatic TCa with high-dose cisplatin-based chemotherapy. [4][5][6][7] However, despite the high survival rate, many TCa patients experience a complex set of psychological, social, and physical morbidity. 8 Chemotherapy-associated acute and long-term physical co-morbidities include nausea and vomiting, gastrointestinal mucositis, fatigue, neurotoxicity, Raynaud-like symptoms, ototoxicity (tinnitus and hearing impairment), thromboembolic events and leukaemia. Radiotherapy long-term toxicity includes radiationinduced secondary malignant neoplasms and decreased sexual satisfaction. 9 In addition, fertility concerns, body image issues and sexual dysfunction (including ejaculatory dysfunction) are intrinsic themes of TCa as the patient group affected are largely men within the fertile age range. 10 Several studies have shown that psychosocial and functional limitations of TCa compromise overall patient QoL and health-related QoL (HR-QoL). These studies have also shown higher rates of depressive symptoms (34%), depression and anxiety disorders (19% vs. 13.5%) compared to the general population. [11][12][13][14] TCa patient-reported outcome measures (PROMs) assessing most prevalent mental health conditions in TCa (anxiety and depression), overall HR-QoL and overall quality of life (QoL) are an integral domain of clinical assessment in TCa person-centred care. 15,16 PROMs measure beyond biological functioning, morbidity, and mortality. They measure the physical, social, and emotional functions of a patient weighted by patient-perceived symptom burden and diagnosis and treatment related expectations. 17 Several validated PROMs of QoL, HR-QoL, anxiety and depression are reported in the clinical assessment of TCa. However, most of these PROMs have been developed as generic tools for a broad spectrum of disorders and domains (e.g., adherence, distress, health-behaviour, functioning) and not specifically for TCa. Importantly no systematic reviews have been identified to psychometrically validate the measurement properties (three domains of reliability, responsiveness, and validity) of these PROMs in a TCa population.
Current PROMs may therefore misrepresent the totality of psychosocial, depressive and anxiety symptoms that compose mental wellbeing and QoL of TCa patients. Clinical assessment performed with PROMs of unknown measurement property quality may further result in missed diagnoses and a greater cumulative psychosocial impact of co-morbidity on QoL in TCa patients and survivors. Therefore, establishing the psychometric properties of these PROMs utilised in TCa is of significant importance in screening, accurate assessment and management of mental health, specifically anxiety and depression, and global psychosocial wellbeing evaluated as QoL.
This systematic review therefore aims to: (i) identify QoL, anxiety and depression PROMs that are psychometrically validated (reliability, validity, and responsiveness) in patients with TCa, (ii) systematically summarise and provide a comprehensive overview of psychometric validation studies and measurement properties of these PROM(s), and (iii) provide recommendations of the most appropriate validated generic and TCa-specific PROM(s).

| Protocol and registration
This systematic review was performed in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Protocols (PRISMA-P) guidelines 18 and the COSMIN methodology for systematic reviews of PROM(s). The review protocol was registered prospectively on PROSPERO, Registration Number: CRD42020160232.

| Study eligibility criteria
Full-text articles that evaluated one or more measurement properties, development of a PROM and/or interpretability of all disease-specific and generic PROMs assessing one of the following principal constructs: QoL, HR-QoL, anxiety and depression, in a study sample of men with TCa (all types, grades and treatment modalities) were included in this review. Measurement properties (quality of PROMs) were defined as per the COnsensus-based Standards for the selection of health status Measurement INstruments) (COSMIN) Taxonomy (Table S1).
The exclusion criteria included: low quality of evidence study designs (case reports, editorials and abstract-only articles); studies with clinician-assessed outcome measurement tools; studies using PROM(s) of interest as an outcome measurement instrument only; validation studies of PROM(s) measuring physical symptoms of wellbeing that is, fertility, erectile dysfunction for selectivity; full-text article not available in the English language.

| Study selection
Titles and abstracts of all articles were assessed for inclusion or exclusion against pre-defined eligibility criteria in duplicate by two independent reviewers (Amine Nur Dincer and Oktay Genel) and disagreements were resolved by discussion between the two reviewers. Included abstracts were retrieved for full-text review and duplicates were identified and deducted. Full-text publications that satisfied the pre-defined study eligibility and exclusion criteria were included. A referencing software (Mendeley 2020, London) was used to manage all citations.   (Table S2). 19 Consistent results of each measurement property quality per PROM were then summarised to produce a total rating per measurement quality per PROM, also recorded as 'sufficient' (+), 'insufficient' (−), or 'indeterminate' (?) and presented in a Summary of Findings (SoF) table (Table 2).

| Data collection and data items
Data extracted on study characteristics included: sample size, participant demographics (age, tumour histological type), country the PROM instrument was administered in, modes of administration (e.g., computer-based vs. paper), response rate (Table 1). Data extracted on PROM(s) included: PROM(s) name; domains/constructs assessed in the PROM(s); item types/response options (such as static or rating scales), scoring system and ranges (Table S3); measurement properties assessed per PROM (Table 2).

| Risk of bias assessment
A risk of bias assessment was performed assessing the methodological quality of included studies using the COSMIN Risk of Bias checklist. 20 The COSMIN Risk of Bias checklist was used to rate the overall quality of individual studies as 'very good', 'adequate', 'doubtful' or 'inadequate'. Heterogeneous terms and definitions for measurement properties in included articles were objectively re-defined to a measurement property defined by the COSMIN taxonomy. 19 The Grading of Recommendations Assessment, Development and Evaluation (GRADE) principles were used to rate the overall quality of evidence of the review findings (Table 2;   Table S4).

| Study selection
A total of 4,305 studies were identified and a further 10 articles were identified through reference screening of full-text publications.
Three hundred and four abstracts satisfied the eligibility criteria or were deemed indeterminate and full-text articles were retrieved and assessed for inclusion. After-deduplication, eight full-text articles were included in this systematic review ( Figure 1).

| Study characteristics and result synthesis
Seven PROM instruments were identified in this review; two of seven  Table 1). Results were synthesised into construct(s) measured and sub-categorised into each measurement property per PROM. Comparative analysis and total PROM score were generated for each measurement property ( Table 2)  Sztankay et al. 21

| Cancer assessment for young adults-testicular
The Cancer Assessment for Young Adults-Testicular (CAYA-T) is a 90-item questionnaire that includes the seven biopsychosocial domains assessing concentration and memory, self-care, education and work, social functioning, sexual relationships, emotional functioning (fear, preoccupation with illness, anxiety and depressive symptoms). One study, with the study population including TCa survivors, psychometrically validating the CAYA-T questionnaire was identified. 24 Hoyt et al. 24  -1425 The lengthiness of the CAYA-T questionnaire may adversely affect response rates, resulting in decreased diagnostic accuracy and reliability. stress/tension items (a = 0.91). Information or model fit on structural validity was not reported and was rated as 'indeterminate'. Internal consistency was also 'indeterminate' as criteria for evidence of 'sufficient' structural validity was not met.   earlier intervention by referral to psychological or psychiatric support or other appropriate services and subsequent better outcomes overall for TCa patients. 36,37 Furthermore, accurate estimation of depression and anxiety prevalence is essential and patient-reported questionnaires report a higher prevalence rate in comparison to clinician-reported questionnaires. 38 However, patient-reported questionnaires are limited in their diagnostic accuracy compared to a standardised diagnostic interview and are therefore best utilised as screening tools. 39 This review also highlights the paucity of primary PROM-validation studies and therefore recommends further validation studies of PROMs assessing anxiety, depression and QoL in TCa patients and survivors that need to be conducted in primary or tertiary clinical settings.

| CONCLUSIONS
The