Outcome measures in palliative care for advanced cancer patients: a review

Transcription

1 Journal of Public Health Medicine Vol. 19, No. 2, pp Printed in Great Britain Outcoe easures in for advanced cancer s: a review Julie Hearn and Irene J. Higginson Suary Inforation generated using outcoe easures to easure the effectiveness of interventions is potentially invaluable. Depending on the easureent tool eployed the results can be used to onitor clinical care, carry out coparative research, provide audit data or infor purchasing decisions. However, the data collected can only ever be as good as the ethod used to obtain the. This review aied to systeatically identify and exaine outcoe easures that have been used, or proposed for use in the clinical audit of of s with advanced cancer. Database searches were perfored using MEDLINE ( ), CANCERLIT ( ), Healthplan ( ), and 'Oncolink' on the Internet. Further easures were located with the assistance of other professionals working in. The criteria for the inclusion and assessent of easures were a easure assessing ore than one doain and a target population of advanced disease or. Fortyone easures were identified, 12 of which satisfied the inclusion criteria. These contained between five and 56 ites and covered aspects of physical, psychological and spiritual doains. Each easure eets soe but not all of the objectives of easureent in, and fulfils soe but not all of our criteria for validity, reliability, responsiveness and appropriateness. Keywords: outcoe easures,, advanced cancer, review. Introduction In the context of health and illness, outcoe is usually defined in ters of the achieveent or failure to achieve desired goals.' The easureent of the health outcoe of interventions can be linked to the assessent of the appropriateness of health care interventions. 2 The use of outcoe easures can therefore help deterine whether a ethod of treatent or particular intervention package is worth while. 3 Consequently, easureent of this 'attributable effect of intervention or its lack on a previous state of health' 4 has iportant iplications for the purchasing of health care services. Outcoe easures in for s with advanced cancer require the easureent of aspects that reflect the specific goals of, such as iproving the quality of life before death, controlling syptos and supporting the faily. 3 Measuring the effectiveness of interventions is becoing increasingly iportant 6 because it allows the evaluation and developent of effective and efficacious teas. A variety of clinical audit tools and systes for palliative care have been developed in recent years, but these are being used in various ways and are constantly changing or being suppleented by new easures. This review aied to identify and exaine outcoe easures that have been used, or proposed for use in the clinical audit of of s with advanced cancer, and to systeatically assess these using well-defined criteria. Method Sources of literature Database searches were perfored using MEDLINE ( ), CANCERLIT ( ), Healthplan ( ), and 'Oncolink' on the Internet (The University of Pennsylvania Cancer Center Resource, ). The search ters used, either singly or in cobination, were audit,, hospice care, terinal care, clinical or edical or nursing audit, quality assurance, audit easures, assessent and outcoe. Further easures were located with the assistance of a ultiprofessional steering group, through personal counications with other professionals working in, and fro an investigation of the grey literature. New easures published during the review period were also identified. Inclusion criteria The criteria for inclusion were: 1. that the target population included cancer s, or s with advanced disease receiving, or was considered by the authors to be appropriate for this group; and Departent of Palliative Care and Policy, King's College School of Medicine, Rayne Institute, Coldharbour Lane, London SE5 9NU. Julie Hearn, Research Assistant Irene J. Higginson, Professor of Palliative Care and Policy Address correspondence to Ms J. Hea. Oxford University Press 1997

2 194 JOURNAL OF PUBLIC HEALTH MEDICINE 2. the easure contained ore than one doain; and 3. the easure could be used on s with all cancer types. Measures that have been used in cancer care but were specific to a particular group, for exaple, leukaeia s (Cancer Leukaeia Group B Studies - CALGB), or easures which concentrated on only one life doain, for exaple, physical syptos (McGill Pain Questionnaire), were excluded fro the review (see Bowling 3 ). It was also iportant to identify only those easures that had been used for s receiving or proposed for use easuring outcoes at this stage of the disease trajectory. For exaple, easures specifically designed to assess the outcoe of nonpalliative cancer cheotherapy, such as the Breast Cancer Cheotherapy Questionnaire (BCCQ), were not included. 3 Assessent of identified easures Measures were then assessed following the criteria outlined in Table 1? Content validity was further assessed by whether the easure covered the particular doainsreportedto berelevantto (physical, psychological and spiritual diensions), and how any ites were contained in each doain. Results In total, 41 easures were identified (see Tables 2 and 3). Twelve of these satisfied the inclusion criteria. These easures contained between five and 56 ites and covered the physical, psychological and spiritual doains of life to differing extents (see Table 3). To suarize the 12 easures: three are copleted by a professional; 6 ' 8 ' 9 seven by the hiself or herself; 10 " 16 two contain both and professional eleents; 17 ' 18 eight assess ites relating only to the, 8 ' 10 " 15 ' 17 whereas four ay also consider the faily or carer unit; 619 ' 1618 seven have been validated in just one setting; 6 ' 10 " ' 1518 five contain 30 or ore ites; 10 ' 12 ' 14 ' two were designed for the assessent of clinical trial interventions This paper will now describe each of these easures in ore detail. An initial assessent of suffering 10 This easure was developed on 259 advanced cancer s in acute hospitals. Afive-pointLikert Scale with scores ranging fro five for 'good' to one for 'bad' was used to record the answers to the 43 questions either by the unaided or by a trained nurse interviewer. The questions have been refined to give a shorter 20-ite questionnaire suitable for use during the initial assessent by a eber of any profession in the hospice or tea. Edonton Sypto Assessent Schedule (ESAS) 11 The ESAS was developed for quick assessent of outcoes in routine practice. This tool consists of nine Visual Analogue Scales (VASs). The draws a ark along a 100 line corresponding to how they feel, with the far left end of the line corresponding to the least degree of syptos, and the far right 'worst' syptos. The ESAS is copleted on adission to hospital and twice daily thereafter by the, or with the assistance of a nurse. Patients who are unable to respond owing to cognitive failure are assessed by their nurse or a specially trained faily eber. The score for each ite is recorded on a bar graph, allowing staff to visualize patterns of sypto control over tie. Further testing of this easure's validity and reliability are required, particularly with reference to the potential bias introduced by a change in the person recording the answers on the VAS as care continues. European Organisation for Research on Cancer Treatent (EORTC QLQ-C30) 12 Developed with lung cancer s to evaluate the quality of life of those s participating in international clinical trials, this self-reporting questionnaire is both a reliable and valid easure of the quality of life of cancer s in research settings. Questions cover the past week and responses are ainly in the forat of a straightforward four-point Likert Scale, ranging fro one for 'not at all' to four for 'very uch'. It contains a generic core with cancer-specific odules and work is being carried out to extend the questionnaire for Table 1 Criteria used to assess outcoe easures 7 Validity - the instruent easures what it intends to easure Content validity - does the easure cover those doains considered iportant? Criterion validity - does the easure correlate with superior easures or predict future outcoe? Construct validity - does the easure confor with the results using other established scales (or discriinate between groups of s)? Reliability -the instruent produces the sae results when repeated on an unchanged population Inter-rater reliability - does the easure produce siilar results when used by different observers? Test-retest reliability - does the easure produce siilar results when used at different points in tie? Internal - do individual ites within the instruent correlate with each other? Responsiveness to change - the instruent is able to detect clinically significant change Has the easure deonstrated change as part of a clinical trial or cohort follow-up? Does the easure discriinate between differing degrees of disease severity? Appropriateness of forat - the instruent is suitable for its intended use

3 PALLIATIVE CARE OUTCOME MEASURES 195 Table 2 Scales used in the assessent of the quality of life of cancer s 3 The World Health Organization (WHO) Functional Scale The Zubrod Scale The Eastern European Cooperative Functional Oncology Perforance Scale (ECOGP) The McGill Pain Questionnaire (MPG) Lasry Sexual Functioning Scale for Breast Cancer Patients WHO Sypto Checklist Medical Research Council (MRC) UK Scale The Qualitator Functional Assessent of Cancer Therapy (FACT-G) Functional Living Index - Cancer (FLIC) Cancer Inventory of Proble Situations (CIPS) Cancer Rehabilitation Evaluation Syste (CARES) Spitzer Quality of Life (QL) Index Linear Analogue Self-Assessent (LASA) Scale Ontario Cancer Institute-Royal Marsden Linear Analogue Self-Assessent Scale Padilla Quality of Life (QL) Scale Multidiensional Quality of Life Scale (MQOLS-CA) Holes and Dickerson Global Quality of Life Scales (Coates) Quality of Life Index Breast Cancer Cheotherapy Questionnaire (BCCQ) Visual Analogue Scale (VAS) for Bone Marrow Transplant Patients European Neuroblastoa Study Group Quality of Life Assessent For - Children (QLAF-C) Cancer Leukaeia Group B Studies (CALGB) Ability Index Burge Quality of Life Severity Scale Ananestic Coparative Self-Assessent (ACSA) WHO Quality of Life Assessent Instruent (WHOQOL) TWiST s with ore advanced cancer. At present, soe questions are thought to be inappropriate for this group and have caused distress in s with advanced disease in a French counity setting (D. LaGabrielle, personal counication, 1995). Hebrew Rehabilitation Centre for Aged Quality of Life (HRCA-QL) 8 Adapted fro the Spitzer Quality of Life Index (a scale developed for doctors to easure the quality of life of their cancer s), with the ite activity being replaced by obility for the older target group. It has not been revalidated and has been criticized for lack of responsiveness in s with advanced disease. Ratings for each ite are scored fro zero to two to give a total score of 0-10 (higher scores equate to a better quality of life). It has been used to evaluate treatents and support services. McGill Quality of Life Questionnaire (MQOL) U Developed on advanced cancer s treated at hoe or in an unit, the MQOL was designed to easure overall quality of life in people with a life-threatening illness and to indicate the areas in which the is doing well or poorly. The circles a nuber on a ten-point categorical scale, with the extrees of least desirable and ost desirable at either end. It includes an existential doain which the authors propose plays a greater role in deterining quality of life in s with local or etastatic disease than in s with no evidence of disease. The McMaster Quality of Life Scale (MQLS) 17 This easure was developed on 83 s to easure the quality of life in a palliative population including cancer s. Ites are rated on a seven-point nuerical scale with the direction of positive and negative descriptors varied. It is currently being refined and s are now asked which ten ites of the scale are ost iportant to their quality of life. Patients who begin to experience difficulty filling in answers are then asked to rate only these ten, ost iportant ites. Palliative Care Assessent (PACA) 6 This easure was developed on 125 s to assess the outcoe of interventions ade within two weeks of referral to a hospital tea. The PACA for coprises three rating scales. Syptos are scored on a four-point scale fro zero for 'absent' to three for 'daily life doinated by the sypto', assessing the severity of each sypto fro the 's perspective, using a sei-structured interview. Insight is assessed by an observer on a five-point scale, and plans for future care were asked of the and recorded

4 Table 3 Measures for assessing the outcoe of for people with advanced cancer Nae of easure (author and year) Nuber of ites and doains covered Validity Reliability Responsiveness to change Appropriateness of forat Setting Tie Adinistration o An Initial Assessent of Suffering 10 Edonton Sypto Assessent Schedule - ESAS 11 European Organisation for Research on Treatent of Cancer - EORTC QLQ-C30 12 Hebrew Rehabilitation Centre for Aged Quality of Life Index - HRCA-QL 8 43 (); ood, syptos, fears and faily worries, knowledge and involveent, support 9 (); pain activity, nausea, depression, anxiety, drowsiness, appetite, wellbeing, shortness of breath 30 (); 9 ulti-ite scales including 5 functional, 3 sypto scales and a global quality of life scale 5 (); obility, daily living, health, attitude, support Spitzer Quality of Life Index physical health groups STAS (except for activity) inter-scale correlation, correlates with clinical status Uni and Multi scale version and with the Kaofsky Perforance Index ( ) ( ) (0 77) and ( ) stable over tie iproveent deonstrated in odule being evaluated in Europe scores correlate with survival out- counity few inutes inutes 1-2 inutes or by professional interview or with nurse assistance professional URNAL OF :PUBLIC HE.A.LTH MEDI CINE The McGill Quality of Life Questionnaire - MQOL' 3 17 (); physical syptos, psychological syptos, outlook on life and eaningful existence Spitzer Quality of Life and SIS (0.89) distinguishes between s out- The McMaster Quality of Life Scale - MQLS (), physical syptos, functional status, social functioning, eotional status, cognition, sleep and rest, energy and vitality, general life satisfaction, eaning of life Spitzer Quality of Life ( ) ( ) changes in scores were related to whether the felt they had changed counity out- s 3-30 inutes, staff under 3 inutes, faily approxiately 3 inutes, faily or staff

5 Palliative Care Assessent - PACA 12 ( and relatives); sypto control, insight and future placeent the sypto scores correlate with the McCorkle sypto distress scale (0 44-1) iproveent deonstrated in few inutes professional Palliative Care Core Standards - PCCS 18 6 core standards and 56 process and outcoe ites ( and carer), sypto control, inforation, currently being tested currently being tested not evaluated as yet expected to take about 10 inutes professional,, carer and the bereaved Rotterda Sypto Checklist - RSCL 14 Support Tea Assessent Schedule - STAS 9 Sypto Distress Scale - SDS 15 The Schedule for the Evaluation of Individual Quality of Life - SEIQoL 16 support, bereaveent care and eotional support, specialist education for staff 34 (); physical and psychosocial syptos 17 ( and carer); pain and sypto control, insight, psychosocial, faily needs, planning affairs, hoe services counication, and support of other professionals 13 (); nausea, ood, loss of appetite, insonia, pain, obility, fatigue, appearance, bowel pattern, concentration 5 doains noinated by the individual; 30 hypothetical scenarios are rated based on these doains and weights are derived for each doain inter-scale correlation for psychological diension, less for physical distress ites 's and faily's ratings and with HRCA-QL global Quality of Life easures McMaster health index questionnaire subscales for health status and physical function ( ) ( ) ( ) test-retest ( ) ( ) ( ) validity ( ) not evaluated iproveent deonstrated in sensitive to changes in treatent over tie does not distinguish between s and controls pre-treatent out- counity hospice counity 8 inutes 2 inutes coplete separate questionnaires professional, in presence of an interviewer as part of a structured interview "0 r r > H n > O c n o2 2 c 73

6 198 JOURNAL OF PUBLIC HEALTH MEDICINE on a four-point scale. Facilitation of the appropriate placeent for hospital s is a fundaental eleent of this easure. Palliative Care Core Standards (PCCS) 18 Originally a set of standards for hospice care and counity teas, this tool has been refined and is currently being piloted in units as separate questionnaires for all those involved with the 's well-being including the professionals, the and the carer. Structure, process, education and training are also covered, resulting in a coprehensive but lengthy tool at present Rotterda Sypto Checklist (RSCL) 14 Developed priarily as a tool to easure the syptos reported by cancer s participating in clinical research, this questionnaire uses a four-point Likert Scale to record responses on the bothersoeness of ites over the last three days or week. Categories range fro 'not at all' through to 'very uch'. The authors suggest it ay be useful in the evaluation of supportive care, but it ay be inappropriate for s to coplete as disease advances. 19 The Support Tea Assessent Schedule (STAS) 9 Developed for use with ultidisciplinary cancer support teas, STAS is a validated easure of the effectiveness of palliative care. 5 Ites were developed by cancer support teas to reflect the goals of. The effect of the ites on the daily life of the over the last week is scored by a professional on a five-point Likert Scale ranging fro zero for 'none' (no effect) up to four for 'overwheling effect'. STAS is widely used in counity settings and has been adapted for use in settings and to assess individual syptos. The Sypto Distress Scale (SDS) 15 This scale was developed for s with a life-threatening disease, either cancer or heart disease, and can be used for all types of cancer. The scale is self adinistered (usually in the presence of an interviewer), withresponsesrated on a five-point Likert Scale ranging fro one for 'no distress' to five for 'extree distress'. It concentrates ainly on the syptos and ood in relation to quality of life. Schedule for the Evaluation of Individual Quality of Life (SEIQoL) 16 This easure was developed fro the technique of 'judgeent analysis' to easure s' level of functioning. The easure allows respondents to noinate the five areas of life which are ost iportant to the, rate their level of functioning or satisfaction with each, and indicate the relative iportance of each area to their total quality of life. It has been tested in a variety of populations and healthy individuals, and has recently been reported for use clinically for s with HTV or AIDS anaged in general practice. Discussion In there are particular concerns about the use and relevance of outcoe easures. The ethod of adinistration of a easure, whether -, professional- or carercopleted, is a priary concern with this population. The advantages and disadvantages of these various ethods of recording inforation have been widely docuented and debated. 3 " 20 In the case of s receiving, there is an inherent difficulty using self- easures as any s are too ill to coplete the, or die early during care. 13 This results in a lack of inforation being recorded, leading to potential bias in the results because those s likely to be experiencing the ost probles are less likely to be included in data collection. As an alternative, a final assessent is soeties copleted by a professional, either before or after death. This affects the validity of a easure designed for by the. Professionally copleted easures are frequently used to overcoe this particular proble, but by their nature cannot accurately reflect how the really feels. Cohen et al. argued that the fact that only half of the population can coplete a questionnaire does not ean that health care professionals should not ask those who can rate their quality of life to do so. 13 The second issue when easuring health outcoe for advanced cancer s is whether a easure includes those doains relevant to ' and does not focus on one aspect alone, be it physical syptos (e.g. the Karnofksy Index 22 ) or the existential doain of self-content and wellbeing. The easures described above address the doains to differing extents, but no single easure covers physical, psychological and spiritual doains in a forat that will provide sufficient or reliable inforation. The purpose of easuring the quality of life and outcoes of the care of s is potentially fourfold. 20 One objective is to obtain ore detailed inforation about the for clinical onitoring to aid and iprove care. A second purpose is to audit the care provided, by deterining whether standards are being achieved and identify potential areas for iproveent. Third, research using outcoe easures to copare services, or to copare care before and after the introduction of a service can be of value in assessing the efficacy of a service, and the cost-effectiveness. Finally, analysis of data generated using outcoe easures can be used to infor purchasers and thereby secure resources for future services. Each of the easures described fulfils the objectives to varying degrees, but none of the easures selected successfully eets all of these, and it is questionable whether any such tool can be developed which will eet all the requireents of an 'ideal tool'. However, there is a need to continue researching and developing outcoe easures in that

7 PALLIATIVE CARE OUTCOME MEASURES 199 address the concerns outlined above and that could easily be ipleented into routine practice. In this way, the provision of can be onitored and we can continue to strive to obtain the best standards of care. Acknowledgeents We would like to thank the NHS Executive Clinical Audit Unit for providing funding for this literature review, which fors part of a larger project developing a new, core easure for use with cancer s receiving in a variety of specialist and non-specialist settings. We would also like to thank all the ebers of the project advisory group for their advice and coents. References 1 Wilkin D, Halla L, Doggett MA Measures of need and outcoe for priary health care. Oxford: Oxford University Press, 1992: 5. 2 Brook RH. Relationship between appropriateness and outcoe. In: Hopkins A, Bostain D, eds. Measuring the outcoes of edical care. London: Royal College of Physicians, Bowling A. Measuring disease: a review of disease-specific quality of life easureent scales. Milton Keynes: Open University, Caian KC. Quality of life in cancer s - a hypothesis. Br J Med Ethics 1984; 10: Higginson IJ, McCarthy M. Validity of the support tea assessent schedule: do staffs' ratings reflect those ade by s or their failies? Palliat Med 1993; 7: Ellershaw JE, Peat SJ, Boys LC. Assessing the effectiveness of a hospital tea. Palliat Med 1995; 9: Rasay M, Winget C, Higginson I. Review: easures to deterine the outcoe of counity services for people with deentia. Age Ageing 1995; 24(1): Morris J, Suissa S, Sherwood S, Greer D. Last days: a study of the quality of life of terinally ill cancer s. J Chron Dis 1986; 39: Higginson I. A counity schedule. In: Higginson I. ed. Clinical audit in. Oxford: Radcliffe Medical Press, 1993: MacAda DB, Sith M. An initial assessent of suffering in terinal illness. Palliat Med 1987; 1: Bruera E, Kuehn N, Miller MJ, Selser P, Macillan K. The Edonton Sypto Assessent Syste (ESAS): a siple ethod for the assessent of s. J Palliat Care 1991; 7(2): Aaronson NK, Ahedzai S, Bergan B, et al. The European Organisation for Research and Treatent of Cancer QLQ-C30: a quality-of-life instruent for use in international clinical trials in oncology. J Nat Cancer Inst 1993; 85: Cohen SR, Mount BM, Strobel MG, Bui F. The McGill Quality of Life Questionnaire: a easure of quality of life appropriate for people with advanced disease. A preliinary study of validity and acceptability. Palliat Med 1995; 9(3): de Haes JCJM, van Knippenberg FCE, Neijt JP. Measuring psychological and physical distress in cancer s: structure and application of the Rotterda Sypto Checklist. Br J Cancer 1990; 62: McCorkle R, Young K. Developent of a sypto distress scale. Cancer Nurs 1978; 101: O'Boyle CA, McGee H, Joyce CRB. Quality of life: assessing the individual. Adv Med Sociol 1994; 5: Sterkenburg CA, Woodward CA. A reliability and validity study of the McMaster Quality of Life Scale (MQLS) for a palliative population. J Palliat Care 1996; 12(1): Trent Hospice Audit Group. Palliative Care Core Standards; a ulti-disciplinary approach. Trent Hospice Audit 1992, c/o Nightingale Macillan Continuing Care Unit, Derby. 19 Rathbone GV, Horsley S, Goacher J. A self-evaluated assessent suitable for seriously ill hospice s. Palliat Med 1994; 8(1): Jenkinson C. Measuring health and edical outcoes. London: UCL Press, Saunders CM. The anageent of terinal disease. London: Edward Arnold, Kaofsky DA, Abelann WH, Craver LF, et al. The use of nitrogen ustards in the palliative treatent of carcinoa. Cancer 1948; I: Accepted on 20 Deceber 1996