Systematic review
DOI: 10.1111/1471-0528.13251 www.bjog.org
Patient-reported outcome measures for use in gynaecological oncology: a systematic review NJ Preston,a N Wilson,b NJ Wood,c J Brine,a J Ferreira,a SG Brearleya a Faculty of Health and Medicine, Lancaster University, Lancaster, UK b Burnley General Hospital, Burnley, UK c Royal Preston Hospital, Preston, UK Correspondence: Dr N Preston, Faculty of Health and Medicine, Lancaster University, Lancaster, LA1 4YT, UK. Email [email protected]
Accepted 4 November 2014. Published Online 5 January 2015.
Background Patient-reported outcome measures (PROMs) are
used to assess the impact of health care on a patient’s health. Within the gynaecological oncology setting, multiple PROMs have been adopted but no assessment has been made in terms of their psychometric qualities and robustness. Objectives To undertake a systematic review to identify the most
psychometrically robust and appropriate PROM used in the gynaecological oncology setting. Search strategy A search of the bibliographic database of the
Oxford PROM group, plus nine additional databases, was carried out along with citation-tracking and hand searches. Selection criteria Studies examining the psychometric properties
of outcome measures tested in gynaecological cancer populations were selected by three blinded reviewers. Data collection and analysis Studies were independently assessed and data extracted. Analysis included an appraisal of the psychometric properties and functionality of the included PROMs to guide recommendations.
Main results Eighteen PROMs tested in gynaecological oncology settings were identified. These were categorised into seven areas of focus, and the most psychometrically robust tools were identified: (1) generic (no recommendation); (2) general cancer (EORTC QLQ-C30 and FACT-G); (3) pelvic cancer (QUEST GY); (4) ovarian cancer (EORTC QLQ-OV28); (5) cervical cancer (EORTC QLQ-CX24); (6) endometrial cancer (EORTC QLQ-EN 24); and (7) vulval cancer (FACT-V). Author’s conclusions Seven PROMs were recommended for use in
six gynaecological populations. No single tool was identified that had been tested in all disease groups. Some showed promise, but a lack of conceptual clarity about the core outcomes and the rationale for use will require further testing using well-constructed studies. Keywords Cancer, gynaecological oncology, metrics, outcomes, patient-reported outcome measures, psychometric properties. Linked article This article is commented on by JH Becker, p. 622 in this issue. To view this mini commentary visit http:// dx.doi.org/10.1111/1471-0528.13265.
Please cite this paper as: Preston NJ, Wilson N, Wood NJ, Brine J, Ferreira J, Brearley SG. Patient-reported outcome measures for use in gynaecological oncology: a systematic review. BJOG 2015;122:615–622.
Introduction Patient-reported outcome measures (PROMs) are increasingly used in medicine and research to assess the impact of health care on a patient’s health.1–4 Unlike other clinical outcome indicators (deaths, adverse incidents, etc.), PROMs provide the patients’ perspective on health, services, and the level of care received, without interpretation by a clinician or anyone else.4,5 A PROM may be generic or diseasespecific, and can assess symptoms, function (physical, emotional, social, and sexual), and health-related quality of life.6,7 Validation of the tool, through testing its psychometric properties within the population in which it is used, is important, as the characteristics are population-specific.
ª 2015 Royal College of Obstetricians and Gynaecologists
There are three main uses of PROMs: clinical trials, supporting individual patient care, and service evaluation.8 There is interest in their use for measuring and benchmarking healthcare service delivery so that the delivery of high-quality health care can be rewarded.9,10 The Department of Health in the UK commissioned a review of PROMs for some cancers, including colorectal, breast, prostate, and lung, the findings of which most commonly recommended the SF36 and EQ5D as generic tools, and the EORTC QLQ C30 or FACT-G as cancer-specific tools.6,11 Within gynaecological oncology multiple PROMs are used,12–15 predominatly as part of clinical trials to demonstrate the equivalence or benefit of novel cancer
615
Preston et al.
treatments.16 The critical assessment of their use as tools to detect disease and implement appropriate treatment also demonstrated patient benefit, however,17 and their completion prior to consultation has been associated with an improvement in the quality and benefit of the consultation.18 To date there has been no assessment of the currently available and used gynaecological oncology tools, and the aim of this study was to undertake a systematic review to identify the most psychometrically robust and appropriate PROM to be used in the gynaecological oncology setting.
Methods A systematic review of the literature using established methodology was undertaken to answer the following question.19 What is the most psychometrically robust and appropriate PROM used in gynaecological oncology setting?
Search strategy The search strategy consisted of two parts. First, an initial search was carried out using a bibliographic database, which was compiled by the Oxford PROM group using a comprehensive search strategy.19 The keywords ‘cancer’ AND ‘gynaecology’ were used to search this database, and individual types of gynaecological cancer were searched for within the free text. Second, as the bibliographic database had not been updated since December 2005, further searching from January 2006 onwards was carried out in the following databases: Medline, AMED, CINAHL, Cochrane Library, Dissertations and Theses (PhD only), EMBASE, Index to Theses (PhD only), PsycINFO, and Web of Science. We used the medical subject headings (MeSH) term ‘Genital neoplasm, female’ AND ‘cancer OR neoplasm* OR malignan*’ AND (gynecol* OR gynaecol*) OR (Endomet* OR vulva*) OR (ovary or ovarian) OR ‘germ cell’ OR (cervix or cervical) OR (vagina* or fallopian tube*) OR (uterine or uterus) OR (trophoblas* or choriocarcionma) to search Medline, and this was adapted for other databases. Where the databases supported ‘cited by’ searches, these were carried out. The search was supplemented by citation-tracking any identified studies and reviewing their reference lists. The following journals were also searched by hand. International Journal of Clinical Oncology Quality of Life Research International Journal of Gynecological Cancer Gynecologic Oncology Journals that had previously been hand-searched to form the Oxford PROM database were searched from 2006 onwards. All other journals were hand-searched from 2001 onwards.
616
Titles and abstracts were reviewed independently by three reviewers to assess if they merited full-text analysis (NW plus SB or NP). Papers focused on patient-reported tools, with evidence of the reliability, validity, or responsiveness of the tool tested in a gynaecological oncology population, and published in English, were included. Papers reporting clinician-assessed instruments, instruments without evidence of psychometric properties, singleitem or narrowly focused tools, review articles, or interim reports superseded by full reports were excluded. Any disagreements were resolved by full-text review. Analysis of the full text was undertaken independently by two reviewers (NW and NP or NW and SB), and articles were excluded if they failed to meet the inclusion criteria. Any disagreement was resolved by a third independent reviewer and discussion.
Data collection and analysis Data were extracted independently, by three reviewers, using an extraction tool based upon the criteria used in previous reviews of the psychometric properties of PROMs.6,19 Information extracted included setting, population studied, tool used, the domains included, and the psychometric properties of the tool.
Psychometric properties Psychometric properties, such as reliability and validity, demonstrate that an instrument consistently measures what it was designed to measure.20 For this review, the psychometric properties of PROM instruments were assessed using data from the original published article, and, where applicable, from other studies undertaking psychometric testing of the instruments, including translations.
Reliability Reliability assesses the extent to which an instrument is free from random error, and its ability to consistently measure an attribute, for example on different occasions or by different observers.6,21 The reliability of a particular measure is dependent upon the context and population studied.21 Two features of reliability were assessed: internal consistency and reproducibility. Validity Validity is the extent to which an instrument measures what it reports to measure. As it is specific to the purpose and setting, it is a continuing process and is not a fixed property of a tool.21 Construct validity examines the relationship of a construct to other sets of variables. The most robust form of assessing convergent and divergent validity is to construct a multitrait–multimethod matrix. Correlation coefficients of >0.5 suggest acceptable convergent validity, and of
ª 2015 Royal College of Obstetricians and Gynaecologists
Systematic review of PROMs in gynaecological oncology
correlation coefficients of 500 patients. The EQ5D had more psychometric testing than the SF36; however, it was arguably limited. There was good evidence of validity for both tools. The lack of testing across disease groups and the limited psychometric evidence in favour of either tool signified that neither could be strongly recommended for use with a mixed gynaecological oncology population.
General cancer Five tools were identified that have been designed for use with patients with any type of cancer: EORTC QLQ-C30, FACT-G, QOL-CS, RSCL, and SRQ. Of these, the two most thoroughly tested tools were the EORTC QLQ C30 and the FACT-G. Both tools had been tested with ovarian and cervical cancer populations, and with similar numbers of patients (over 1000 for each tool). Similar psychometric properties had been assessed and were found to have good evidence or some good evidence in terms of reliability, construct validity, and acceptability. Both tools covered a broad range of domains. As a result, both the FACT-G and the EORTC QLQ-C30 are recommended as the most robust and appropriate general cancer tools for use within gynaecological oncology patients.
617
Preston et al.
Table 1. Appraisal of psychometric properties and functionality of the identified PROMs Outcome measure EORTC QLQ-C30 EORTC QLQ-OV28 EORTC QLQ-CX24 EORTC QLQ-EN24 EQ-5D FACT-G FACT-O FACT-V FACT-CX LENT-SOMA NFOSI-18 QOL-CS QuEST GY RSCL SCNS-G SF36 SRQ UGQ Key: 0, not reported; favour.
Reliability
Construct validity
Content validity
Criterion validity
Responsiveness
Acceptability
++ ++ +++ +++ +++ +++ ++ ++ +++ + ++ ++ ++ ++ +++ +++ ++ +++
++ +++ +++ +++ + +++ +++ +++ + 0 +++ + +++ + +++ 0 + 0
0 ++ +++ +++ 0 0 +++ +++ 0 0 +++ 0 +++ 0 0 0 + +++
+ +++ + + ++ ++ +++ +++ 0 0 +++ 0 0 0 0 0 ++ 0
++ + 0 0 0 0 +++ +++ 0 + 0 0 0 0 0 0 0 0
++ + ++ ++ 0 + +
0 + 0
, no evidence in favour; +, some limited evidence in favour; ++, some good evidence in favour; +++, good evidence in
Pelvic cancer Four tools were identified as being designed for use and assessed within a population with cancer affecting the pelvic region: QUEST GY, SCNS, UGQ, and LENT SOMA. QUEST GY and SCNS were found to be the most psychometrically robust, with good evidence or some good evidence in favour of both; however, QUEST GY was developed with more comprehensive testing, and in addition was marginally shorter than the SCNS, and also had been developed to be used with a touch screen, and as such is recommended as the most robust and appropriate tool for use in pelvic cancer populations.
Ovarian cancer Three tools designed to be used for patients with ovarian cancer were identified: EORTC QLQ-OV28, FACT-O, and NFOSI-18. Of these, both EORTC QLQ-OV28 and FACTO were comprehensively psychometrically tested and found to have good evidence or some good evidence in favour, with greater evidence of content validity and responsiveness in FACT-O. Overall, however, the EORTC QLQ-OV28 was tested in more studies using more settings and twice the total number of patients, and as a result is recommended as the most robust tool and appropriate tool for use with ovarian cancer patients.
Cervical cancer Two tools were identified that were designed and tested for use in cervical cancer populations: FACT-CX and EORTC
618
0 + +++ 0 ++ ++
QLQ-CX24. Both tools demonstrated robust psychometric properties in terms of reliability. The EORTC tool was also tested for construct, content, and criterion validity and acceptability, and showed some evidence to robust evidence in favour. In addition, these studies were undertaken in a diverse range of settings, and with the largest population of cervical cancer patients, and therefore the EORTC QLQCX24 is recommended as the most appropriate and robust tool for use with this population.
Endometrial cancer The EORTC QLQ-EN24 was the single tool identified that was designed for use with endometrial cancer patients. It was found to have good evidence in its favour in terms of reliability, construct validity, content validity, and acceptability, and some limited evidence in its favour with criterion validity. As such, this tool can be recommended for use within the endometrial cancer population.
Vulval cancer The review identified one tool that was developed and tested within patients with cancer of the vulva: FACT-V. It was found to be psychometrically robust, with some good evidence in favour in terms of reliability, and good evidence when assessed for construct, content, and criterion validity and responsiveness. Therefore, FACT-V is recommended as the most psychometrically robust and appropriate tool for use within the vulval cancer population.
ª 2015 Royal College of Obstetricians and Gynaecologists
Systematic review of PROMs in gynaecological oncology
Discussion Main findings The aim of this review was to assess PROMs tested within a gynaecological oncology setting, and to make an assessment of their psychometric robustness. In practice the nature of gynaecological oncology, with its different tumour types, disease trajectories, and concomitant symptoms, means that identifying a single PROM for use across all gynaecological tumour types is challenging. In part this is because of a lack of conceptual clarity about what constitutes the core outcomes within this population, but also there are issues about the scope and length of the tool, and the rationale for administering it. The scope of the instrument encompasses the domains it covers, and the associated breadth and depth of its items, which ranged from assessing 3 to 11 domains in the identified instruments (Table S2). This, however, does not reflect how comprehensively each domain was assessed: for example, the EQ5D assesses six domains and comprises six items, which indicates a single item assessment of each domain, signifying that it is relatively broad in scope but that the assessment is not of great depth. In addition, the lack of clarity regarding the key outcomes for measurement within gynaecological oncology populations signifies that it is not clear which constitute core domains within a PROM. As such, any assessment of scope is limited to how comprehensive the tool is rather than its application to the population, which may mean important disease-specific outcomes are missed. For example, the EORTC QLQ-C30 and FACT-G are broad in scope but do not address sexual function, which is arguably a central issue with women with gynaecological cancers. Finally, there is considerable variation between the domains, which can encompass function, wellbeing, or specific symptoms, and this lack of equal weighting means that they are difficult to quantify. The tools identified varied considerably in length, domains covered, and scoring. The variation in the number of items, which ranged from 6 to 66 (Table S2), has implications for both the scope of the tool and the time for completion, with its associated patient burden. What constitutes an acceptable length of time or level of burden depends on performance status, which signifies that completion time may be less important within a follow-up clinic of cancer survivors than with patients with advanced disease who are unwell. There are diverse motivations for administering PROMs within a clinical setting, which include research, service evaluation, or to support individualised patient care. As such, attention must be paid to the rationale for adopting a PROM in order to assess if the instrument selected is robust, appropriate, and fits into clinical realities. There
ª 2015 Royal College of Obstetricians and Gynaecologists
needs to be agreement about the purpose of a PROM within the gynaecological oncology setting, and whether it will be used as a systematic method of assessing practice (benchmarking), or for individualised patient-centred care, or both. The different uses of PROMs have resulted in a range of instruments being developed (18 in this review alone). These use a variety of scoring mechanisms, which has implications in terms of access to scores: for example, tools such as the EORTC QLQ use a scoring algorithm, which arguably precludes their use for individualised care. Other tools use a summed score mechanism, such as the FACT tools, or average scores across domains, such as the QOLCS, and may have potentially greater versatility in their use. A number of the tools lacked clarity in the scoring mechanism adopted, and require clarification. In selecting a tool for clinical practice that may have been developed for research, attention must be given to how and when scores will be accessed, and also to practical considerations such as the length of staff time needed in terms of training, administration, and scoring. The review identified a number of PROMs that have been tested with mixed gynaecological cancer patient groups and have shown some promising results. The most widely used PROM across gynaecological cancer groups is the EORTC QLQ-C30.16 This gives a broad assessment of patients’ experiences, but for clinicians or researchers wanting to assess domains that might be particular to a specific cancer site, a site-specific tool will be useful. The QUEST GY is a relatively new addition to this field, and therefore there is less evidence of its use than with more established PROMs such as the EORTC QLQ-C30; however, it may represent a useful tool in assessing cancers originating in the pelvic area. No tool has been tested across all disease groups within gynaecological oncology. Moreover, not all tools have been tested in multiple settings and countries, which may have implications in terms of their appropriateness for use outside these settings.
Strengths and limitations The main strength of this research is the rigorous and transparent approach adopted to locate and extract data from papers examining PROMs in gynaecological cancer patient populations. It has identified tools within seven broad areas of focus within these. It has generated clear recommendations on the most psychometrically robust and appropriate tools for use within general and site-specific gynaecological oncology. The information identified is only as good as that provided within the included papers. Not all psychometric testing may have been detailed, and information on
619
Preston et al.
acceptability, responsiveness, and the scoring system was often omitted from articles. A number of tools were relatively new, which may have implications on the number of studies in which they were tested and the concomitant size and heterogeneity of the population, and/or the reporting of psychometric properties. This may have resulted in a bias towards older and more tested tools. The study adopted the appraisal system of the Oxford PROM group to assess the quality of the psychometric properties and functionality of the PROMs (Table 1).19 In order to minimise the subjectivity of this simple scoring system, three reviewers independently scored each tool and consensus was required for the final agreed rating. Lastly, the review excluded papers not published in English.
Contribution to authorship
Interpretations
None.
The review has identified a number of promising tools; however, none were tested within all populations within gynaecological oncology. Currently there is a lack of conceptual clarity about the core outcomes for measurement within this population. There also needs to be clarity on the purpose of the tool: benchmarking or individualised patient care. Future research should bring together a panel of experts and patients and use consensus techniques to develop the recommendations of this research, and decide whether an existing tool can be adopted, whether an existing tool should be developed further, or whether attention should be given to developing a new PROM. This might require further testing across the tumour types: for example, QUEST GY has good coverage of domains and promising psychometric properties, but is currently supported by only a single study.
Conclusion This review aimed to identify the most psychometrically robust and appropriate tools for use within gynaecological oncology, resulting in seven PROMs recommended for use across six broad populations. No single tool could be recommended for all populations, in part because of a lack of conceptual clarity about the core outcomes requiring testing and the rationale for using PROMs (individualised care or benchmarking). Some tools showed promise but the identification of a single PROM for use within gynaecologic oncology will need to be demonstrated through further well-constructed studies following standard PROM methodology.
Disclosure of interests No interests to disclose.
620
The study was conceived by NJP, SB, and NJW. The design of the study was undertaken by NJP, SB, NW, and JB. Data were collected by NJP, SB, NW, JB, and NJW. Interpretations of the data were undertaken by NJP, SB, NW, NJW, and JFA. Drafting of the article and final approval was the responsibility of all authors.
Details of ethics approval No ethical approval was required as this was a systematic review of published papers.
Funding Financial support for this study was provided by the Rosemere Cancer Foundation, through an unrestricted grant.
Acknowledgements
Supporting Information Additional Supporting Information may be found in the online version of this article: Figure S1. PRISMA flow diagram showing a selection of papers for the review. Table S1. Psychometric properties of the outcome measures. Table S2. Structure of the tools and domains covered. Table S3. Summary of recommendations of PROMs by population. &
References 1 Aaronson NK, Snyder C. Using patient-reported outcomes in clinical practice: proceedings of an International Society of Quality of Life Research conference. Qual Life Res 2008;17:1295. 2 McKenna SP, Doward LC. Integrating patient-reported outcomes. Value Health 2004;7(Suppl 1):S9–12. 3 Snyder C, Brundage M. Integrating patient-reported outcomes in healthcare policy, research and practice. Expert Rev Pharmacoecon Outcomes Res 2010;10:351–3. 4 Food and Drug Administration. Guidance for industry: patient-reported outcome measures: use in medical product development to support labeling claims. 2009 [www.fda.gov/downloads/Drugs/Guidances/ UCM193282.pdf] Accessed 27 October 2014. 5 Devlin NJ, Appleby J. Getting the Most Out of Proms: Putting Health Outcomes at the Heart of NHS Decision-Making. London: The King’s Fund, 2010 [www.kingsfund.org.uk/sites/files/kf/Getting-the-mostout-of-PROMs-Nancy-Devlin-John-Appleby-Kings-Fund-March-2010.pdf]. Accessed 20 October 2014. 6 Fitzpatrick R, Davey C, Buxton MJ, Jones DR. Evaluating patientbased outcome measures for use in clinical trials. Health Technol Assess 1998;2:i–iv, 1–74. 7 Valderas JM, Alonso J. Patient reported outcome measures: a model-based classification system for research and clinical practice. Qual Life Res 2008;17:1125–35.
ª 2015 Royal College of Obstetricians and Gynaecologists
Systematic review of PROMs in gynaecological oncology
8 Lipscomb J, Donaldson MS, Hiatt RA. Cancer outcomes research and the arenas of application. J Natl Cancer Inst Monogr 2004; 1–7. 9 Health & Social Information Centre. Provisional monthly patient reported outcome measures (PROMs) in England, 2013 [www.hscic.gov.uk/media/1537/A-Guide-to-PROMs-Methodology/pdf/ PROMS_Guide_v5.pdf]. Accessed 20 October 2014. 10 United Kingdom Parliament. Improving quality in the NHS, 2009 [www.publications.parliament.uk/pa/cm200809/cmselect/cmhealth/ 53/5307.htm]. Accessed 11 July 2013. 11 Patient reported outcome measurement group. Home page University of Oxford. [http://phi.uhce.ox.ac.uk/home.php]. Accessed 11 July 2013. 12 Greimel E, Nordin A, Lanceley A, Creutzberg CL, van de Poll-Franse LV, Radisic VB, et al. Psychometric validation of the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire-Endometrial Cancer Module (EORTC QLQ-EN24). Eur J Cancer 2011;47:183–90. 13 Jensen SE, Rosenbloom SK, Beaumont JL, Abernethy A, Jacobsen PB, Syrjala K, et al. A new index of priority symptoms in advanced ovarian cancer. Gynecol Oncol 2011;120:214–9. 14 Colagiuri B, King MT, Butow PN, McGrane JA, Luckett T, Price MA, et al. A comparison of the FACT-G and the Supportive Care Needs Survey (SCNS) in women with ovarian cancer: unidimensionality of constructs. Qual Life Res 2012;21:887–97. 15 Harley C, Takeuchi E, Taylor S, Keding A, Absolom K, Brown J, et al. A mixed methods approach to adapting health-related quality of life measures for use in routine oncology clinical practice. Qual Life Res 2012;21:389–403. 16 Jones GL, Ledger W, Bonnett TJ, Radley S, Parkinson N, Kennedy SH. The impact of treatment for gynecological cancer on healthrelated quality of life (HRQoL): a systematic review. Am J Obstet Gynecol 2006;194:26–42. 17 Marshall S, Haywood K, Fitzpatrick R. Impact of patient-reported outcome measures on routine practice: a structured review. J Eval Clin Pract 2006;12:559–68. 18 Velikova G, Booth L, Smith AB, Brown PM, Lynch P, Brown JM, et al. Measuring quality of life in routine oncology practice improves communication and patient well-being: a randomized controlled trial. J Clin Oncol 2004;22:714–24. 19 Gibbons E, Fitzpatrick R. A Structured Review of PatientReported Outcome Measures for People Undergoing Elective Procedures for Benign Prostatic Hyperplasia. Oxford: University of Oxford, 2011. 20 Brearley SG, Clements CV, Molassiotis A. A review of patient selfreport tools for chemotherapy-induced nausea and vomiting. Support Care Cancer 2008;16:1213–29. 21 Streiner DL, Norman GR. Health Measurement Scales: A Practical Guide to Their Development and Use, 3rd edn. Oxford: Oxford University Press, 2003. 22 Campbell DT, Fiske DW. Convergent and discriminant validation by the multitrait-multimethod matrix. Psychol Bull 1959;56:81–105. 23 Guyatt GH, Feeny DH, Patrick DL. Measuring health-related quality of life. Ann Intern Med 1993;118:622–9. 24 Liang MH. Longitudinal construct validity: establishment of clinical meaning in patient evaluative instruments. Med Care 2000;38(9 Suppl):II84–90. 25 Ashing-Giwa KT, Kim J, Tejero JS. Measuring quality of life among cervical cancer survivors: preliminary assessment of instrumentation validity in a cross-cultural study. Qual Life Res 2008;17:147–57. 26 Lang HC, Chuang L, Shun SC, Hsieh CL, Lan CF. Validation of EQ5D in patients with cervical cancer in Taiwan. Support Care Cancer 2010;18:1279–86.
ª 2015 Royal College of Obstetricians and Gynaecologists
27 Osoba D, Zee B, Pater J, Warr D, Kaizer L, Latreille J. Psychometric properties and responsiveness of the EORTC quality of Life Questionnaire (QLQ-C30) in patients with breast, ovarian and lung cancer. Qual Life Res 1994;3:353–64. 28 Zhao H, Kanda K. Translation and validation of the standard Chinese version of the EORTC QLQ-C30. Qual Life Res 2000;9:129– 37. 29 Park SY, Bae DS, Nam JH, Park CT, Cho CH, Lee JM, et al. Quality of life and sexual problems in disease-free survivors of cervical cancer compared with the general population. Cancer 2007;110:2716–25. 30 Ersek M, Ferrell BR, Dow KH, Melancon CH. Quality of life in women with ovarian cancer. West J Nurs Res 1997;19: 334–50. 31 Wenzel LB, Donnelly JP, Fowler JM, Habbal R, Taylor TH, Aziz N, et al. Resilience, reflection, and residual stress in ovarian cancer survivorship: a gynecologic oncology group study. Psychooncology 2002;11:142–53. 32 van de Poll-Franse LV, Mols F, Essink-Bot ML, Haartsen JE, Vingerhoets AJ, Lybeert ML, et al. Impact of external beam adjuvant radiotherapy on health-related quality of life for longterm survivors of endometrial adenocarcinoma: a population-based study. Int J Radiat Oncol Biol Phys 2007;69:125–32. 33 Reis N, Beji NK, Coskun A. Quality of life and sexual functioning in gynecological cancer patients: results from quantitative and qualitative data. Eur J Oncol Nurs 2010;14:137–46. 34 de Haes JC, van Knippenberg FC, Neijt JP. Measuring psychological and physical distress in cancer patients: structure and application of the Rotterdam Symptom Checklist. Br J Cancer 1990;62:1034–8. 35 Donovan HS, Ward S, Sherwood P, Serlin RC. Evaluation of the Symptom Representation Questionnaire (SRQ) for assessing cancerrelated symptoms. J Pain Symptom Manage 2008;35:242–57. 36 Jensen PT, Klee MC, Groenvold M. Validation of a questionnaire for self-rating of urological and gynaecological morbidity after treatment of gynaecological cancer. Radiother Oncol 2002;65:29– 38. 37 Routledge JA, Burns MP, Swindell R, Khoo VS, West CML, Davidson SE. Evaluation of the lent-soma scales for the prospective assessment of treatment morbidity in cervical carcinoma. Int J Radiat Oncol Biol Phys 2003;56:502–10. 38 Steele R, Fitch MI. Supportive care needs of women with gynecologic cancer. Cancer Nurs 2008;31:284–91. 39 Greimel E, Bottomley A, Cull A, Waldenstrom AC, Arraras J, Chauvenet L, et al. An international field study of the reliability and validity of a disease-specific questionnaire module (the QLQ-OV28) in assessing the quality of life of patients with ovarian cancer. Eur J Cancer 2003;39:1402–8. 40 Chie WC, Lan CY, Chiang C, Chen CA. Quality of life of patients with ovarian cancer in Taiwan: validation and application of the Taiwan Chinese version of the EORTC QLQ-OV28. Psychooncology 2010;19:782–5. 41 Basen-Engquist K, Bodurka-Bevers D, Fitzgerad MA, Webster K, Cella D, Hu SH, et al. Reliability and validity of the functional assessment of cancer therapy-ovarian. J Clin Oncol 2001;19:1809– 17. 42 Janda M, Obermair A, Cella D, Perrin LC, Nicklin JL, Ward BG, et al. The functional assessment of cancer-vulvar: reliability and validity. Gynecol Oncol 2005;97:568–75. 43 Greimel ER, Kuljanic Vlasic K, Waldenstrom AC, Duric VM, Jensen PT, Singer S, et al. The European Organization for Research
621
Preston et al.
and Treatment of Cancer (EORTC) Quality-of-Life questionnaire cervical cancer module: EORTC QLQ-CX24. Cancer 2006;107:1812– 22. 44 Jayasekara H, Rajapaksa LC, Greimel ER. The EORTC QLQ-CX24 cervical cancer-specific quality of life questionnaire: psychometric properties in a South Asian sample of cervical cancer patients. Psychooncology 2008;17:1053–7.
45 Singer S, Kuhnt S, Momenghalibaf A, Stuhr C, Dimmel-Hennersdorf U, Kohler U, et al. Patients’ acceptance and psychometric properties of the EORTC QLQ-CX24 after surgery. Gynecol Oncol 2010; 116:82–7. 46 Fernandes WC, Kimura M. Health related quality of life of women with cervical cancer. Rev Lat Am Enfermagem 2010;18:360–7.
Capturing outcomes: doctor doesn’t always know best JH Becker Department of obstetrics and gynaecology, Zuwe Hofpoort Hospital, Woerden, the Netherlands Linked article: This is a mini commentary on NJ Preston et al., pp. 615–22 in this issue. To view this article visit http://dx.doi.org/10.1111/1471-0528.13251. Published Online 5 January 2015. Patient report outcome measures, or PROMs, have been around for a while, but they have gained importance in recent years. In the past, doctors always had the tendency to act somewhat parentally: the doctor knows what’s best for you. But do they really? Insight into patients’ views of their illness and treatment has recently gained attention and importance. Not only do individual doctors and hospital management have an increasing interest in this, but insurance companies do as well and they are dictating physicians to measure PROMs more and more. We see this in obstetrics and gynaecology. One of the problems of PROMs, however, is that the measured outcomes are not as definite as the more traditional outcome measures, such as death, survival, duration of surgery or blood loss. It is therefore of utmost importance that we try to come to an integral, uniform set of patient-reported outcomes: after all, this will allow studies to be reproduced and to be compared with each other. Preston et al. have done a
622
wonderful job by assessing different PROMs used in gynaecological oncology. For most areas of focus they come to a recommendation of specific PROMs to use, but for some areas of focus these recommendations could not be made [Preston et al. BJOG 2015; DOI: 10.1111/ 1471-0528.13251]. Very appropriately, they state that we must pay attention to the rationale for adopting PROMs and must select those PROMs that best fit the purpose of our research setting. Harmonising outcomes reporting in women’s health research is one of the goals of the CROWN initiative [BJOG 2014;121:1181–2] (www.crown-initiative.org). This initiative, in which many international peer-reviewed journals are involved, aims to encourage researchers to report core outcome sets for key conditions in women’s health. If future research could report these core outcome sets, traditional outcomes as well as patient-reported outcomes, future studies could be better compared to each other and be better combined in possible meta-analyses.
Therefore, let me emphasise the authors’ recommendation: future research should bring together a panel of experts and patients and use consensus techniques to develop the recommendations of this research and decide whether an existing tool can be adopted, whether an existing tool should be developed further or whether attention should be given to developing a new PROM. To get the most out of clinical research, researchers ought to think about including follow-up studies when setting up a randomised controlled trial, containing well-defined PROMs. After all, it is the patient who can tell the doctor which of the treatment strategies has worked best for their physical and/or mental wellbeing. Don’t just ask what you can do for the patient, but ask yourself what the patient can do for you as well!
Disclosure of interest The author has no conflicts of interest to declare and no funding to declare. &
ª 2015 Royal College of Obstetricians and Gynaecologists
Copyright of BJOG: An International Journal of Obstetrics & Gynaecology is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
DOI: 10.1111/1471-0528.13251 www.bjog.org
Patient-reported outcome measures for use in gynaecological oncology: a systematic review NJ Preston,a N Wilson,b NJ Wood,c J Brine,a J Ferreira,a SG Brearleya a Faculty of Health and Medicine, Lancaster University, Lancaster, UK b Burnley General Hospital, Burnley, UK c Royal Preston Hospital, Preston, UK Correspondence: Dr N Preston, Faculty of Health and Medicine, Lancaster University, Lancaster, LA1 4YT, UK. Email [email protected]
Accepted 4 November 2014. Published Online 5 January 2015.
Background Patient-reported outcome measures (PROMs) are
used to assess the impact of health care on a patient’s health. Within the gynaecological oncology setting, multiple PROMs have been adopted but no assessment has been made in terms of their psychometric qualities and robustness. Objectives To undertake a systematic review to identify the most
psychometrically robust and appropriate PROM used in the gynaecological oncology setting. Search strategy A search of the bibliographic database of the
Oxford PROM group, plus nine additional databases, was carried out along with citation-tracking and hand searches. Selection criteria Studies examining the psychometric properties
of outcome measures tested in gynaecological cancer populations were selected by three blinded reviewers. Data collection and analysis Studies were independently assessed and data extracted. Analysis included an appraisal of the psychometric properties and functionality of the included PROMs to guide recommendations.
Main results Eighteen PROMs tested in gynaecological oncology settings were identified. These were categorised into seven areas of focus, and the most psychometrically robust tools were identified: (1) generic (no recommendation); (2) general cancer (EORTC QLQ-C30 and FACT-G); (3) pelvic cancer (QUEST GY); (4) ovarian cancer (EORTC QLQ-OV28); (5) cervical cancer (EORTC QLQ-CX24); (6) endometrial cancer (EORTC QLQ-EN 24); and (7) vulval cancer (FACT-V). Author’s conclusions Seven PROMs were recommended for use in
six gynaecological populations. No single tool was identified that had been tested in all disease groups. Some showed promise, but a lack of conceptual clarity about the core outcomes and the rationale for use will require further testing using well-constructed studies. Keywords Cancer, gynaecological oncology, metrics, outcomes, patient-reported outcome measures, psychometric properties. Linked article This article is commented on by JH Becker, p. 622 in this issue. To view this mini commentary visit http:// dx.doi.org/10.1111/1471-0528.13265.
Please cite this paper as: Preston NJ, Wilson N, Wood NJ, Brine J, Ferreira J, Brearley SG. Patient-reported outcome measures for use in gynaecological oncology: a systematic review. BJOG 2015;122:615–622.
Introduction Patient-reported outcome measures (PROMs) are increasingly used in medicine and research to assess the impact of health care on a patient’s health.1–4 Unlike other clinical outcome indicators (deaths, adverse incidents, etc.), PROMs provide the patients’ perspective on health, services, and the level of care received, without interpretation by a clinician or anyone else.4,5 A PROM may be generic or diseasespecific, and can assess symptoms, function (physical, emotional, social, and sexual), and health-related quality of life.6,7 Validation of the tool, through testing its psychometric properties within the population in which it is used, is important, as the characteristics are population-specific.
ª 2015 Royal College of Obstetricians and Gynaecologists
There are three main uses of PROMs: clinical trials, supporting individual patient care, and service evaluation.8 There is interest in their use for measuring and benchmarking healthcare service delivery so that the delivery of high-quality health care can be rewarded.9,10 The Department of Health in the UK commissioned a review of PROMs for some cancers, including colorectal, breast, prostate, and lung, the findings of which most commonly recommended the SF36 and EQ5D as generic tools, and the EORTC QLQ C30 or FACT-G as cancer-specific tools.6,11 Within gynaecological oncology multiple PROMs are used,12–15 predominatly as part of clinical trials to demonstrate the equivalence or benefit of novel cancer
615
Preston et al.
treatments.16 The critical assessment of their use as tools to detect disease and implement appropriate treatment also demonstrated patient benefit, however,17 and their completion prior to consultation has been associated with an improvement in the quality and benefit of the consultation.18 To date there has been no assessment of the currently available and used gynaecological oncology tools, and the aim of this study was to undertake a systematic review to identify the most psychometrically robust and appropriate PROM to be used in the gynaecological oncology setting.
Methods A systematic review of the literature using established methodology was undertaken to answer the following question.19 What is the most psychometrically robust and appropriate PROM used in gynaecological oncology setting?
Search strategy The search strategy consisted of two parts. First, an initial search was carried out using a bibliographic database, which was compiled by the Oxford PROM group using a comprehensive search strategy.19 The keywords ‘cancer’ AND ‘gynaecology’ were used to search this database, and individual types of gynaecological cancer were searched for within the free text. Second, as the bibliographic database had not been updated since December 2005, further searching from January 2006 onwards was carried out in the following databases: Medline, AMED, CINAHL, Cochrane Library, Dissertations and Theses (PhD only), EMBASE, Index to Theses (PhD only), PsycINFO, and Web of Science. We used the medical subject headings (MeSH) term ‘Genital neoplasm, female’ AND ‘cancer OR neoplasm* OR malignan*’ AND (gynecol* OR gynaecol*) OR (Endomet* OR vulva*) OR (ovary or ovarian) OR ‘germ cell’ OR (cervix or cervical) OR (vagina* or fallopian tube*) OR (uterine or uterus) OR (trophoblas* or choriocarcionma) to search Medline, and this was adapted for other databases. Where the databases supported ‘cited by’ searches, these were carried out. The search was supplemented by citation-tracking any identified studies and reviewing their reference lists. The following journals were also searched by hand. International Journal of Clinical Oncology Quality of Life Research International Journal of Gynecological Cancer Gynecologic Oncology Journals that had previously been hand-searched to form the Oxford PROM database were searched from 2006 onwards. All other journals were hand-searched from 2001 onwards.
616
Titles and abstracts were reviewed independently by three reviewers to assess if they merited full-text analysis (NW plus SB or NP). Papers focused on patient-reported tools, with evidence of the reliability, validity, or responsiveness of the tool tested in a gynaecological oncology population, and published in English, were included. Papers reporting clinician-assessed instruments, instruments without evidence of psychometric properties, singleitem or narrowly focused tools, review articles, or interim reports superseded by full reports were excluded. Any disagreements were resolved by full-text review. Analysis of the full text was undertaken independently by two reviewers (NW and NP or NW and SB), and articles were excluded if they failed to meet the inclusion criteria. Any disagreement was resolved by a third independent reviewer and discussion.
Data collection and analysis Data were extracted independently, by three reviewers, using an extraction tool based upon the criteria used in previous reviews of the psychometric properties of PROMs.6,19 Information extracted included setting, population studied, tool used, the domains included, and the psychometric properties of the tool.
Psychometric properties Psychometric properties, such as reliability and validity, demonstrate that an instrument consistently measures what it was designed to measure.20 For this review, the psychometric properties of PROM instruments were assessed using data from the original published article, and, where applicable, from other studies undertaking psychometric testing of the instruments, including translations.
Reliability Reliability assesses the extent to which an instrument is free from random error, and its ability to consistently measure an attribute, for example on different occasions or by different observers.6,21 The reliability of a particular measure is dependent upon the context and population studied.21 Two features of reliability were assessed: internal consistency and reproducibility. Validity Validity is the extent to which an instrument measures what it reports to measure. As it is specific to the purpose and setting, it is a continuing process and is not a fixed property of a tool.21 Construct validity examines the relationship of a construct to other sets of variables. The most robust form of assessing convergent and divergent validity is to construct a multitrait–multimethod matrix. Correlation coefficients of >0.5 suggest acceptable convergent validity, and of
ª 2015 Royal College of Obstetricians and Gynaecologists
Systematic review of PROMs in gynaecological oncology
correlation coefficients of 500 patients. The EQ5D had more psychometric testing than the SF36; however, it was arguably limited. There was good evidence of validity for both tools. The lack of testing across disease groups and the limited psychometric evidence in favour of either tool signified that neither could be strongly recommended for use with a mixed gynaecological oncology population.
General cancer Five tools were identified that have been designed for use with patients with any type of cancer: EORTC QLQ-C30, FACT-G, QOL-CS, RSCL, and SRQ. Of these, the two most thoroughly tested tools were the EORTC QLQ C30 and the FACT-G. Both tools had been tested with ovarian and cervical cancer populations, and with similar numbers of patients (over 1000 for each tool). Similar psychometric properties had been assessed and were found to have good evidence or some good evidence in terms of reliability, construct validity, and acceptability. Both tools covered a broad range of domains. As a result, both the FACT-G and the EORTC QLQ-C30 are recommended as the most robust and appropriate general cancer tools for use within gynaecological oncology patients.
617
Preston et al.
Table 1. Appraisal of psychometric properties and functionality of the identified PROMs Outcome measure EORTC QLQ-C30 EORTC QLQ-OV28 EORTC QLQ-CX24 EORTC QLQ-EN24 EQ-5D FACT-G FACT-O FACT-V FACT-CX LENT-SOMA NFOSI-18 QOL-CS QuEST GY RSCL SCNS-G SF36 SRQ UGQ Key: 0, not reported; favour.
Reliability
Construct validity
Content validity
Criterion validity
Responsiveness
Acceptability
++ ++ +++ +++ +++ +++ ++ ++ +++ + ++ ++ ++ ++ +++ +++ ++ +++
++ +++ +++ +++ + +++ +++ +++ + 0 +++ + +++ + +++ 0 + 0
0 ++ +++ +++ 0 0 +++ +++ 0 0 +++ 0 +++ 0 0 0 + +++
+ +++ + + ++ ++ +++ +++ 0 0 +++ 0 0 0 0 0 ++ 0
++ + 0 0 0 0 +++ +++ 0 + 0 0 0 0 0 0 0 0
++ + ++ ++ 0 + +
0 + 0
, no evidence in favour; +, some limited evidence in favour; ++, some good evidence in favour; +++, good evidence in
Pelvic cancer Four tools were identified as being designed for use and assessed within a population with cancer affecting the pelvic region: QUEST GY, SCNS, UGQ, and LENT SOMA. QUEST GY and SCNS were found to be the most psychometrically robust, with good evidence or some good evidence in favour of both; however, QUEST GY was developed with more comprehensive testing, and in addition was marginally shorter than the SCNS, and also had been developed to be used with a touch screen, and as such is recommended as the most robust and appropriate tool for use in pelvic cancer populations.
Ovarian cancer Three tools designed to be used for patients with ovarian cancer were identified: EORTC QLQ-OV28, FACT-O, and NFOSI-18. Of these, both EORTC QLQ-OV28 and FACTO were comprehensively psychometrically tested and found to have good evidence or some good evidence in favour, with greater evidence of content validity and responsiveness in FACT-O. Overall, however, the EORTC QLQ-OV28 was tested in more studies using more settings and twice the total number of patients, and as a result is recommended as the most robust tool and appropriate tool for use with ovarian cancer patients.
Cervical cancer Two tools were identified that were designed and tested for use in cervical cancer populations: FACT-CX and EORTC
618
0 + +++ 0 ++ ++
QLQ-CX24. Both tools demonstrated robust psychometric properties in terms of reliability. The EORTC tool was also tested for construct, content, and criterion validity and acceptability, and showed some evidence to robust evidence in favour. In addition, these studies were undertaken in a diverse range of settings, and with the largest population of cervical cancer patients, and therefore the EORTC QLQCX24 is recommended as the most appropriate and robust tool for use with this population.
Endometrial cancer The EORTC QLQ-EN24 was the single tool identified that was designed for use with endometrial cancer patients. It was found to have good evidence in its favour in terms of reliability, construct validity, content validity, and acceptability, and some limited evidence in its favour with criterion validity. As such, this tool can be recommended for use within the endometrial cancer population.
Vulval cancer The review identified one tool that was developed and tested within patients with cancer of the vulva: FACT-V. It was found to be psychometrically robust, with some good evidence in favour in terms of reliability, and good evidence when assessed for construct, content, and criterion validity and responsiveness. Therefore, FACT-V is recommended as the most psychometrically robust and appropriate tool for use within the vulval cancer population.
ª 2015 Royal College of Obstetricians and Gynaecologists
Systematic review of PROMs in gynaecological oncology
Discussion Main findings The aim of this review was to assess PROMs tested within a gynaecological oncology setting, and to make an assessment of their psychometric robustness. In practice the nature of gynaecological oncology, with its different tumour types, disease trajectories, and concomitant symptoms, means that identifying a single PROM for use across all gynaecological tumour types is challenging. In part this is because of a lack of conceptual clarity about what constitutes the core outcomes within this population, but also there are issues about the scope and length of the tool, and the rationale for administering it. The scope of the instrument encompasses the domains it covers, and the associated breadth and depth of its items, which ranged from assessing 3 to 11 domains in the identified instruments (Table S2). This, however, does not reflect how comprehensively each domain was assessed: for example, the EQ5D assesses six domains and comprises six items, which indicates a single item assessment of each domain, signifying that it is relatively broad in scope but that the assessment is not of great depth. In addition, the lack of clarity regarding the key outcomes for measurement within gynaecological oncology populations signifies that it is not clear which constitute core domains within a PROM. As such, any assessment of scope is limited to how comprehensive the tool is rather than its application to the population, which may mean important disease-specific outcomes are missed. For example, the EORTC QLQ-C30 and FACT-G are broad in scope but do not address sexual function, which is arguably a central issue with women with gynaecological cancers. Finally, there is considerable variation between the domains, which can encompass function, wellbeing, or specific symptoms, and this lack of equal weighting means that they are difficult to quantify. The tools identified varied considerably in length, domains covered, and scoring. The variation in the number of items, which ranged from 6 to 66 (Table S2), has implications for both the scope of the tool and the time for completion, with its associated patient burden. What constitutes an acceptable length of time or level of burden depends on performance status, which signifies that completion time may be less important within a follow-up clinic of cancer survivors than with patients with advanced disease who are unwell. There are diverse motivations for administering PROMs within a clinical setting, which include research, service evaluation, or to support individualised patient care. As such, attention must be paid to the rationale for adopting a PROM in order to assess if the instrument selected is robust, appropriate, and fits into clinical realities. There
ª 2015 Royal College of Obstetricians and Gynaecologists
needs to be agreement about the purpose of a PROM within the gynaecological oncology setting, and whether it will be used as a systematic method of assessing practice (benchmarking), or for individualised patient-centred care, or both. The different uses of PROMs have resulted in a range of instruments being developed (18 in this review alone). These use a variety of scoring mechanisms, which has implications in terms of access to scores: for example, tools such as the EORTC QLQ use a scoring algorithm, which arguably precludes their use for individualised care. Other tools use a summed score mechanism, such as the FACT tools, or average scores across domains, such as the QOLCS, and may have potentially greater versatility in their use. A number of the tools lacked clarity in the scoring mechanism adopted, and require clarification. In selecting a tool for clinical practice that may have been developed for research, attention must be given to how and when scores will be accessed, and also to practical considerations such as the length of staff time needed in terms of training, administration, and scoring. The review identified a number of PROMs that have been tested with mixed gynaecological cancer patient groups and have shown some promising results. The most widely used PROM across gynaecological cancer groups is the EORTC QLQ-C30.16 This gives a broad assessment of patients’ experiences, but for clinicians or researchers wanting to assess domains that might be particular to a specific cancer site, a site-specific tool will be useful. The QUEST GY is a relatively new addition to this field, and therefore there is less evidence of its use than with more established PROMs such as the EORTC QLQ-C30; however, it may represent a useful tool in assessing cancers originating in the pelvic area. No tool has been tested across all disease groups within gynaecological oncology. Moreover, not all tools have been tested in multiple settings and countries, which may have implications in terms of their appropriateness for use outside these settings.
Strengths and limitations The main strength of this research is the rigorous and transparent approach adopted to locate and extract data from papers examining PROMs in gynaecological cancer patient populations. It has identified tools within seven broad areas of focus within these. It has generated clear recommendations on the most psychometrically robust and appropriate tools for use within general and site-specific gynaecological oncology. The information identified is only as good as that provided within the included papers. Not all psychometric testing may have been detailed, and information on
619
Preston et al.
acceptability, responsiveness, and the scoring system was often omitted from articles. A number of tools were relatively new, which may have implications on the number of studies in which they were tested and the concomitant size and heterogeneity of the population, and/or the reporting of psychometric properties. This may have resulted in a bias towards older and more tested tools. The study adopted the appraisal system of the Oxford PROM group to assess the quality of the psychometric properties and functionality of the PROMs (Table 1).19 In order to minimise the subjectivity of this simple scoring system, three reviewers independently scored each tool and consensus was required for the final agreed rating. Lastly, the review excluded papers not published in English.
Contribution to authorship
Interpretations
None.
The review has identified a number of promising tools; however, none were tested within all populations within gynaecological oncology. Currently there is a lack of conceptual clarity about the core outcomes for measurement within this population. There also needs to be clarity on the purpose of the tool: benchmarking or individualised patient care. Future research should bring together a panel of experts and patients and use consensus techniques to develop the recommendations of this research, and decide whether an existing tool can be adopted, whether an existing tool should be developed further, or whether attention should be given to developing a new PROM. This might require further testing across the tumour types: for example, QUEST GY has good coverage of domains and promising psychometric properties, but is currently supported by only a single study.
Conclusion This review aimed to identify the most psychometrically robust and appropriate tools for use within gynaecological oncology, resulting in seven PROMs recommended for use across six broad populations. No single tool could be recommended for all populations, in part because of a lack of conceptual clarity about the core outcomes requiring testing and the rationale for using PROMs (individualised care or benchmarking). Some tools showed promise but the identification of a single PROM for use within gynaecologic oncology will need to be demonstrated through further well-constructed studies following standard PROM methodology.
Disclosure of interests No interests to disclose.
620
The study was conceived by NJP, SB, and NJW. The design of the study was undertaken by NJP, SB, NW, and JB. Data were collected by NJP, SB, NW, JB, and NJW. Interpretations of the data were undertaken by NJP, SB, NW, NJW, and JFA. Drafting of the article and final approval was the responsibility of all authors.
Details of ethics approval No ethical approval was required as this was a systematic review of published papers.
Funding Financial support for this study was provided by the Rosemere Cancer Foundation, through an unrestricted grant.
Acknowledgements
Supporting Information Additional Supporting Information may be found in the online version of this article: Figure S1. PRISMA flow diagram showing a selection of papers for the review. Table S1. Psychometric properties of the outcome measures. Table S2. Structure of the tools and domains covered. Table S3. Summary of recommendations of PROMs by population. &
References 1 Aaronson NK, Snyder C. Using patient-reported outcomes in clinical practice: proceedings of an International Society of Quality of Life Research conference. Qual Life Res 2008;17:1295. 2 McKenna SP, Doward LC. Integrating patient-reported outcomes. Value Health 2004;7(Suppl 1):S9–12. 3 Snyder C, Brundage M. Integrating patient-reported outcomes in healthcare policy, research and practice. Expert Rev Pharmacoecon Outcomes Res 2010;10:351–3. 4 Food and Drug Administration. Guidance for industry: patient-reported outcome measures: use in medical product development to support labeling claims. 2009 [www.fda.gov/downloads/Drugs/Guidances/ UCM193282.pdf] Accessed 27 October 2014. 5 Devlin NJ, Appleby J. Getting the Most Out of Proms: Putting Health Outcomes at the Heart of NHS Decision-Making. London: The King’s Fund, 2010 [www.kingsfund.org.uk/sites/files/kf/Getting-the-mostout-of-PROMs-Nancy-Devlin-John-Appleby-Kings-Fund-March-2010.pdf]. Accessed 20 October 2014. 6 Fitzpatrick R, Davey C, Buxton MJ, Jones DR. Evaluating patientbased outcome measures for use in clinical trials. Health Technol Assess 1998;2:i–iv, 1–74. 7 Valderas JM, Alonso J. Patient reported outcome measures: a model-based classification system for research and clinical practice. Qual Life Res 2008;17:1125–35.
ª 2015 Royal College of Obstetricians and Gynaecologists
Systematic review of PROMs in gynaecological oncology
8 Lipscomb J, Donaldson MS, Hiatt RA. Cancer outcomes research and the arenas of application. J Natl Cancer Inst Monogr 2004; 1–7. 9 Health & Social Information Centre. Provisional monthly patient reported outcome measures (PROMs) in England, 2013 [www.hscic.gov.uk/media/1537/A-Guide-to-PROMs-Methodology/pdf/ PROMS_Guide_v5.pdf]. Accessed 20 October 2014. 10 United Kingdom Parliament. Improving quality in the NHS, 2009 [www.publications.parliament.uk/pa/cm200809/cmselect/cmhealth/ 53/5307.htm]. Accessed 11 July 2013. 11 Patient reported outcome measurement group. Home page University of Oxford. [http://phi.uhce.ox.ac.uk/home.php]. Accessed 11 July 2013. 12 Greimel E, Nordin A, Lanceley A, Creutzberg CL, van de Poll-Franse LV, Radisic VB, et al. Psychometric validation of the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire-Endometrial Cancer Module (EORTC QLQ-EN24). Eur J Cancer 2011;47:183–90. 13 Jensen SE, Rosenbloom SK, Beaumont JL, Abernethy A, Jacobsen PB, Syrjala K, et al. A new index of priority symptoms in advanced ovarian cancer. Gynecol Oncol 2011;120:214–9. 14 Colagiuri B, King MT, Butow PN, McGrane JA, Luckett T, Price MA, et al. A comparison of the FACT-G and the Supportive Care Needs Survey (SCNS) in women with ovarian cancer: unidimensionality of constructs. Qual Life Res 2012;21:887–97. 15 Harley C, Takeuchi E, Taylor S, Keding A, Absolom K, Brown J, et al. A mixed methods approach to adapting health-related quality of life measures for use in routine oncology clinical practice. Qual Life Res 2012;21:389–403. 16 Jones GL, Ledger W, Bonnett TJ, Radley S, Parkinson N, Kennedy SH. The impact of treatment for gynecological cancer on healthrelated quality of life (HRQoL): a systematic review. Am J Obstet Gynecol 2006;194:26–42. 17 Marshall S, Haywood K, Fitzpatrick R. Impact of patient-reported outcome measures on routine practice: a structured review. J Eval Clin Pract 2006;12:559–68. 18 Velikova G, Booth L, Smith AB, Brown PM, Lynch P, Brown JM, et al. Measuring quality of life in routine oncology practice improves communication and patient well-being: a randomized controlled trial. J Clin Oncol 2004;22:714–24. 19 Gibbons E, Fitzpatrick R. A Structured Review of PatientReported Outcome Measures for People Undergoing Elective Procedures for Benign Prostatic Hyperplasia. Oxford: University of Oxford, 2011. 20 Brearley SG, Clements CV, Molassiotis A. A review of patient selfreport tools for chemotherapy-induced nausea and vomiting. Support Care Cancer 2008;16:1213–29. 21 Streiner DL, Norman GR. Health Measurement Scales: A Practical Guide to Their Development and Use, 3rd edn. Oxford: Oxford University Press, 2003. 22 Campbell DT, Fiske DW. Convergent and discriminant validation by the multitrait-multimethod matrix. Psychol Bull 1959;56:81–105. 23 Guyatt GH, Feeny DH, Patrick DL. Measuring health-related quality of life. Ann Intern Med 1993;118:622–9. 24 Liang MH. Longitudinal construct validity: establishment of clinical meaning in patient evaluative instruments. Med Care 2000;38(9 Suppl):II84–90. 25 Ashing-Giwa KT, Kim J, Tejero JS. Measuring quality of life among cervical cancer survivors: preliminary assessment of instrumentation validity in a cross-cultural study. Qual Life Res 2008;17:147–57. 26 Lang HC, Chuang L, Shun SC, Hsieh CL, Lan CF. Validation of EQ5D in patients with cervical cancer in Taiwan. Support Care Cancer 2010;18:1279–86.
ª 2015 Royal College of Obstetricians and Gynaecologists
27 Osoba D, Zee B, Pater J, Warr D, Kaizer L, Latreille J. Psychometric properties and responsiveness of the EORTC quality of Life Questionnaire (QLQ-C30) in patients with breast, ovarian and lung cancer. Qual Life Res 1994;3:353–64. 28 Zhao H, Kanda K. Translation and validation of the standard Chinese version of the EORTC QLQ-C30. Qual Life Res 2000;9:129– 37. 29 Park SY, Bae DS, Nam JH, Park CT, Cho CH, Lee JM, et al. Quality of life and sexual problems in disease-free survivors of cervical cancer compared with the general population. Cancer 2007;110:2716–25. 30 Ersek M, Ferrell BR, Dow KH, Melancon CH. Quality of life in women with ovarian cancer. West J Nurs Res 1997;19: 334–50. 31 Wenzel LB, Donnelly JP, Fowler JM, Habbal R, Taylor TH, Aziz N, et al. Resilience, reflection, and residual stress in ovarian cancer survivorship: a gynecologic oncology group study. Psychooncology 2002;11:142–53. 32 van de Poll-Franse LV, Mols F, Essink-Bot ML, Haartsen JE, Vingerhoets AJ, Lybeert ML, et al. Impact of external beam adjuvant radiotherapy on health-related quality of life for longterm survivors of endometrial adenocarcinoma: a population-based study. Int J Radiat Oncol Biol Phys 2007;69:125–32. 33 Reis N, Beji NK, Coskun A. Quality of life and sexual functioning in gynecological cancer patients: results from quantitative and qualitative data. Eur J Oncol Nurs 2010;14:137–46. 34 de Haes JC, van Knippenberg FC, Neijt JP. Measuring psychological and physical distress in cancer patients: structure and application of the Rotterdam Symptom Checklist. Br J Cancer 1990;62:1034–8. 35 Donovan HS, Ward S, Sherwood P, Serlin RC. Evaluation of the Symptom Representation Questionnaire (SRQ) for assessing cancerrelated symptoms. J Pain Symptom Manage 2008;35:242–57. 36 Jensen PT, Klee MC, Groenvold M. Validation of a questionnaire for self-rating of urological and gynaecological morbidity after treatment of gynaecological cancer. Radiother Oncol 2002;65:29– 38. 37 Routledge JA, Burns MP, Swindell R, Khoo VS, West CML, Davidson SE. Evaluation of the lent-soma scales for the prospective assessment of treatment morbidity in cervical carcinoma. Int J Radiat Oncol Biol Phys 2003;56:502–10. 38 Steele R, Fitch MI. Supportive care needs of women with gynecologic cancer. Cancer Nurs 2008;31:284–91. 39 Greimel E, Bottomley A, Cull A, Waldenstrom AC, Arraras J, Chauvenet L, et al. An international field study of the reliability and validity of a disease-specific questionnaire module (the QLQ-OV28) in assessing the quality of life of patients with ovarian cancer. Eur J Cancer 2003;39:1402–8. 40 Chie WC, Lan CY, Chiang C, Chen CA. Quality of life of patients with ovarian cancer in Taiwan: validation and application of the Taiwan Chinese version of the EORTC QLQ-OV28. Psychooncology 2010;19:782–5. 41 Basen-Engquist K, Bodurka-Bevers D, Fitzgerad MA, Webster K, Cella D, Hu SH, et al. Reliability and validity of the functional assessment of cancer therapy-ovarian. J Clin Oncol 2001;19:1809– 17. 42 Janda M, Obermair A, Cella D, Perrin LC, Nicklin JL, Ward BG, et al. The functional assessment of cancer-vulvar: reliability and validity. Gynecol Oncol 2005;97:568–75. 43 Greimel ER, Kuljanic Vlasic K, Waldenstrom AC, Duric VM, Jensen PT, Singer S, et al. The European Organization for Research
621
Preston et al.
and Treatment of Cancer (EORTC) Quality-of-Life questionnaire cervical cancer module: EORTC QLQ-CX24. Cancer 2006;107:1812– 22. 44 Jayasekara H, Rajapaksa LC, Greimel ER. The EORTC QLQ-CX24 cervical cancer-specific quality of life questionnaire: psychometric properties in a South Asian sample of cervical cancer patients. Psychooncology 2008;17:1053–7.
45 Singer S, Kuhnt S, Momenghalibaf A, Stuhr C, Dimmel-Hennersdorf U, Kohler U, et al. Patients’ acceptance and psychometric properties of the EORTC QLQ-CX24 after surgery. Gynecol Oncol 2010; 116:82–7. 46 Fernandes WC, Kimura M. Health related quality of life of women with cervical cancer. Rev Lat Am Enfermagem 2010;18:360–7.
Capturing outcomes: doctor doesn’t always know best JH Becker Department of obstetrics and gynaecology, Zuwe Hofpoort Hospital, Woerden, the Netherlands Linked article: This is a mini commentary on NJ Preston et al., pp. 615–22 in this issue. To view this article visit http://dx.doi.org/10.1111/1471-0528.13251. Published Online 5 January 2015. Patient report outcome measures, or PROMs, have been around for a while, but they have gained importance in recent years. In the past, doctors always had the tendency to act somewhat parentally: the doctor knows what’s best for you. But do they really? Insight into patients’ views of their illness and treatment has recently gained attention and importance. Not only do individual doctors and hospital management have an increasing interest in this, but insurance companies do as well and they are dictating physicians to measure PROMs more and more. We see this in obstetrics and gynaecology. One of the problems of PROMs, however, is that the measured outcomes are not as definite as the more traditional outcome measures, such as death, survival, duration of surgery or blood loss. It is therefore of utmost importance that we try to come to an integral, uniform set of patient-reported outcomes: after all, this will allow studies to be reproduced and to be compared with each other. Preston et al. have done a
622
wonderful job by assessing different PROMs used in gynaecological oncology. For most areas of focus they come to a recommendation of specific PROMs to use, but for some areas of focus these recommendations could not be made [Preston et al. BJOG 2015; DOI: 10.1111/ 1471-0528.13251]. Very appropriately, they state that we must pay attention to the rationale for adopting PROMs and must select those PROMs that best fit the purpose of our research setting. Harmonising outcomes reporting in women’s health research is one of the goals of the CROWN initiative [BJOG 2014;121:1181–2] (www.crown-initiative.org). This initiative, in which many international peer-reviewed journals are involved, aims to encourage researchers to report core outcome sets for key conditions in women’s health. If future research could report these core outcome sets, traditional outcomes as well as patient-reported outcomes, future studies could be better compared to each other and be better combined in possible meta-analyses.
Therefore, let me emphasise the authors’ recommendation: future research should bring together a panel of experts and patients and use consensus techniques to develop the recommendations of this research and decide whether an existing tool can be adopted, whether an existing tool should be developed further or whether attention should be given to developing a new PROM. To get the most out of clinical research, researchers ought to think about including follow-up studies when setting up a randomised controlled trial, containing well-defined PROMs. After all, it is the patient who can tell the doctor which of the treatment strategies has worked best for their physical and/or mental wellbeing. Don’t just ask what you can do for the patient, but ask yourself what the patient can do for you as well!
Disclosure of interest The author has no conflicts of interest to declare and no funding to declare. &
ª 2015 Royal College of Obstetricians and Gynaecologists
Copyright of BJOG: An International Journal of Obstetrics & Gynaecology is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
