Generic Preference Research Articles

Getting specific: participation preference in urooncological decision-making

BackgroundShared decision-making is the gold standard for good clinical practice, and thus, psychometric instruments have been established to assess patients’ generic preference for participation (e.g., the Autonomy Preference Index, API). However, patients’ preferences may vary depending on the specific disease and with respect to the specific decision context. With a modified preference index (API-Uro), we assessed patients’ specific participation preference in preference-sensitive decisions pertaining to urological cancer treatments and compared this with their generic participation preference.MethodsIn Study 1, we recruited (N = 469) urological outpatients (43.1% urooncological) at a large university hospital. Participation preference was assessed with generic measures (API and API case vignettes) and with the disease-specific API-Uro (urooncological case vignettes describing medical decisions of variable difficulty). A polychoric exploratory factor analysis was used to establish factorial validity and reduce items.In Study 2, we collected data from N = 204 bladder cancer patients in a multicenter study to validate the factorial structure with confirmatory factor analysis.Differences between the participation preference for different decision contexts were analyzed.ResultsStudy 1: Scores on the specific urooncological case vignettes (API-Uro) correlated with the generic measure (r = .44) but also provided incremental information. Among the disease-specific vignettes of the API-Uro, there were two factors with good internal consistency (α ≥ .8): treatment versus diagnostic decisions. Patients desired more participation for treatment decisions (77.8%) than for diagnostic decisions (22%), χ2(1) = 245.1, p ≤ .001.Study 2: Replicated the correlation of the API-Uro with the API (r = .39) and its factorial structure (SRMR = .08; CFI = .974). Bladder cancer patients also desired more participation for treatment decisions (57.4%) than for diagnostic decisions (13.3%), χ²(1) =84, p ≤ .001.ConclusionsThe desire to participate varies between treatment versus diagnostic decisions among urological patients. This underscores the importance of assessing participation preference for specific contexts. Overall, the new API-Uro has good psychometric properties and is well suited to assess patients’ preferences. In routine care, measures of participation preference for specific decision contexts may provide incremental, allowing clinicians to better address their patients’ individual needs.

Assessing the validity and responsiveness of a generic preference quality of life measure in the context of posttraumatic stress disorder

PurposeThere is limited research exploring the usefulness of generic preference-based quality of life (GPQoL) measures used to facilitate economic evaluation in the context of posttraumatic stress disorder (PTSD). The aim of the current study was to explore the validity and responsiveness of a common GPQoL measure (Assessment of Quality of Life 8 Dimension [AQoL-8D]) in relation to a PTSD condition-specific outcome measure (Posttraumatic Stress Disorder Checklist for the DSM-5 [PCL-5]).MethodThis aim was investigated in a sample of individuals (N = 147) who received trauma-focused cognitive-behavioural therapies for posttraumatic stress disorder. Convergent validity was investigated using spearman’s correlations, and the level of agreement was investigated using Bland–Altman plots. Responsiveness was investigated by exploring the standardised response means (SRM) from pre-post-treatment across the two measures, which allow the comparison of the magnitude of change between the measures over time.ResultsCorrelations between the AQoL-8D (dimensions, utility and summary total scores) and the PCL-5 total score ranged from small to large and agreement between the measures was considered moderate to good. While SRMs were large for the AQoL-8D and PCL-5 total scores, the SRM for the PCL-5 was nearly double that of the AQoL-8D.ConclusionOur findings demonstrate that the AQoL-8D has good construct validity but present preliminary evidence that economic evaluations using only GPQoL measures may not fully capture the effectiveness of PTSD treatments.

PMH29 ESTIMATING A PREFERENCE-BASE INDEX FROM THE RECOVERING QUALITY OF LIFE MEASURE FOR ECONOMIC EVALUATION IN THE AREA OF MENTAL HEALTH: VALUATION OF REQOL-UI

There are increasing concerns about the use of a generic preference based measure to capture health benefits in the area of mental health. We present the preference weights for a new preference-based measure for mental health, the ReQoL-UI, derived from Recovering Quality of Life (ReQoL-10), a recovery measure developed with mental health service users. Item response theory analyses were undertaken using survey responses from a sample of mental health service users (n = 4266) to derive a health state classification system from ReQoL-10 and to select health states amenable to valuation. A valuation survey with 305 members of the UK general population representative in terms of age, gender and region was conducted using face-to-face interviewer administered time-trade-off (TTO) with props. To produce utility values for every health state described by the ReQoL-UI a series of regression models including random effects models were fitted to the data. The ReQoL-UI health state classification system comprises six mental health themes (activity; belonging; choice, control and autonomy; hope; self-perception and well-being) and one physical health item, with 5 severity levels for each item. Sixty-four health states were valued. The selected model was a random effects model involving squared terms and interactions between mental health and the physical health item. The model had significant and consistent coefficients, RMSE of 0.07 and lowest Akaike and Bayesian Information Criterion. ReQoL-UI has been generated using a novel method to select health state values and widely accepted interviewer administered TTO to elicit preferences. The valuation of the ReQoL-UI produced a valid set of preference weights that can be used to generate QALYs for use in cost-utility analysis of interventions in the area of mental health.

Quality of life in older adults following a hip fracture: an empirical comparison of the ICECAP-O and the EQ-5D-3\xa0L instruments

BackgroundThe purpose of this study was to empirically compare the performance of two generic preference based quality of life instruments, EQ-5D-3 L (with a health and physical function focus) and ICECAP-O (with a wellbeing and capability focus), in a population of older Australians following hip fracture.MethodsOlder adults or their family member proxies (in cases of severe cognitive impairment) following surgery to repair a fractured hip were invited to take part in this cross sectional study. Inclusion criteria included an age of 60 years or older, confirmed falls-related hip fracture and those receiving current palliative care or consented to other research studies were excluded. 87 participants completed the ICECAP-O and EQ-5D-3 L instruments between one and three weeks post-surgery.ResultsFor the hip fracture population, the mean ICECAP-O score was 0.639 (SD 0.206, n = 82) and the mean EQ-5D-3 L utility score was 0.545 (SD 0.251, n = 87). There was a statistically significant positive correlation between the ICECAP-O and EQ-5D-3 L scores (r = 0.529, p = < 0.001).ConclusionsStudy findings indicate significant impairments in quality of life post hip fracture. In multiple regression analyses, age and health-related QoL (measured by the EQ-5D) were significant determinants of ICECAP-O scores, while proxy respondent status, age, and capability-related QoL (measured by the ICECAP-O) were significant determinants of EQ-5D scores. Approaches to measuring and valuing quality of life in this sample, should consider the target domains of the intervention in their choice of instrument. Studies aiming to measure the impact of interventions targeting broader domains of wellbeing and QoL should consider including a broader measure of QoL in conjunction with a HRQoL measure.

Generic Reimbursed Market Shares In Greece From 2012-2016. A Study Based On Age Group Preferences

A generic drug shares the same active and inactive ingredients as the branded though it is marketed under a different trade name, trademark, packaging than the branded. Increased penetration of prescribed generics has been primary goal during the memorandum years (since 2010) in Greece. We assessed trends in the prescriptions of authorized generics in the public reimburesed market from EOPYY in Greece during the years 2012-2016. Medicines Data were derived from the EOF (Greek Medicines Organization) databases, the Greek Positive list of prescribed medicines and published reimbursement data by EOPYY (Greek Public Reimbursement Organization) .Time trends in generic market were described and compared with chi-square tests in each semester from 2012 to 2016. In 2012 generics market share was 22% both in values and volume of the public reimbursed market. In 2016 Gx value and volume shares were approximately 25%. Generic value share increased linearly from 72% in S2 2012 to 80% in S2 2015 in patients age group >60 years old. In S1&S2 2016 Generic value share decreased significantly to 72% in patient group >60 years old but increased in almost all other age groups. In the same period older patients switched from generics to Off patent medicines in a rate 10% per semester and to On patent medicines in a rate 20% per semester. Chronic deseases patients count for over 70% volume in public reimbursed market shares. Generic medicines are preferred by older patients. The use of generics in reimbursed market is increased overtime. The effect of increased generic penetration on the reimbursed market and the incentives for generic preference should be assessed.

Inductive Pairwise Ranking: Going Beyond the &lt;i&gt;n&lt;/i&gt; log(&lt;i&gt;n&lt;/i&gt;) Barrier

We study the problem of ranking a set of items from nonactively chosen pairwise preferences where each item has feature information with it. We propose and characterize a very broad class of preference matrices giving rise to the Feature Low Rank (FLR) model, which subsumes several models ranging from the classic Bradley–Terry–Luce (BTL) (Bradley and Terry 1952) and Thurstone (Thurstone 1927) models to the recently proposed blade-chest (Chen and Joachims 2016) and generic low-rank preference (Rajkumar and Agarwal 2016) models. We use the technique of matrix completion in the presence of side information to develop the Inductive Pairwise Ranking (IPR) algorithm that provably learns a good ranking under the FLR model, in a sample-efficient manner. In practice, through systematic synthetic simulations, we confirm our theoretical findings regarding improvements in the sample complexity due to the use of feature information. Moreover, on popular real-world preference learning datasets, with as less as 10% sampling of the pairwise comparisons, our method recovers a good ranking.

Environmental notification support using passage detecting sensor nodes

Interconnected sensor nodes can be used for a long term monitoring of widely spread ecosystems. Individual nodes can be unavailable during the runtime due to a lack of energy, disruption of communication channels, or physical damages of nodes. In this study, a new approach to transfer messages from nodes to gateways (and optional to flood gate nodes) in such dynamically disturbed networks is presented. The focus is the routing in branched line-structures, for example, flood barriers. The usage of gradient structures in each node allow tree routing, tree propagation, energy-efficient exploring broadcast, exploitation of line-structures, and collection of position information. In addition, a generic preference evaluation and decision making concept allows energy or signal quality evaluation, routing decision, structure setup decision, position estimation, synchronisation adaption, and randomisation of security approaches. It can be extended by further rules.

Valuing the Child Health Utility 9D: Using profile case best worst scaling methods to develop a new adolescent specific scoring algorithm

In contrast to the recent proliferation of studies incorporating ordinal methods to generate health state values from adults, to date relatively few studies have utilised ordinal methods to generate health state values from adolescents. This paper reports upon a study to apply profile case best worst scaling methods to derive a new adolescent specific scoring algorithm for the Child Health Utility 9D (CHU9D), a generic preference based instrument that has been specifically designed for the estimation of quality adjusted life years for the economic evaluation of health care treatment and preventive programs targeted at young people. A survey was developed for administration in an on-line format in which consenting community based Australian adolescents aged 11–17 years (N = 1982) indicated the best and worst features of a series of 10 health states derived from the CHU9D descriptive system. The data were analyzed using latent class conditional logit models to estimate values (part worth utilities) for each level of the nine attributes relating to the CHU9D. A marginal utility matrix was then estimated to generate an adolescent-specific scoring algorithm on the full health = 1 and dead = 0 scale required for the calculation of QALYs. It was evident that different decision processes were being used in the best and worst choices. Whilst respondents appeared readily able to choose ‘best’ attribute levels for the CHU9D health states, a large amount of random variability and indeed different decision rules were evident for the choice of ‘worst’ attribute levels, to the extent that the best and worst data should not be pooled from the statistical perspective. The optimal adolescent-specific scoring algorithm was therefore derived using data obtained from the best choices only. The study provides important insights into the use of profile case best worst scaling methods to generate health state values with adolescent populations.

Preferences for Prostate Cancer Outcomes: A Comparison of the Patient Perspective, the General Population Perspective, and a Population at Risk for Prostate Cancer

ObjectiveTo collect disease-specific and generic preference values for three populations. MethodsProstate cancer-specific health states were developed with attributes that varied across five health domains: sexual function, urinary function, bowel function, pain, and fear of the future. Men with prostate cancer, men at risk for prostate cancer, and a sample of the general population assigned value to 18 disease-specific health states using standard gamble (SG) methodology. Study participants also completed the Health Utilities Index (HUI) to obtain generic, community-based preference values to capture their current health rating. ResultsA total of 136 participants were enrolled (n = 43 prostate cancer; n = 40 at risk for prostate cancer; n = 49 general population). Mean HUI mark 3 current health ratings: men with prostate cancer 0.75 ± 0.260; men at risk for prostate cancer 0.77 ± 0.238; general population 0.84 ± 0.178. Mean SG preference values ranged from 0.46 to 0.85 among men with prostate cancer, 0.37 to 0.75 among men at risk for prostate cancer, and 0.32 to 0.81 among the general population group. ConclusionsIn general, preference values for disease-specific health states using the patient perspective were higher than those for the general population. Generic preference values calculated from the HUI were higher than disease-specific preference values calculated from the SG. The higher values calculated from the HUI, from all three perspectives, indicate that a generic measure may not be sensitive enough to capture the disutility of prostate cancer symptoms, specifically sexual dysfunction, urinary dysfunction, and bowel dysfunction, which are being directly measured in the disease-specific health states.

A systematic review of instruments for measuring outcomes in economic evaluation within aged care.

BackgroundThis paper describes the methods and results of a systematic review to identify instruments used to measure quality of life outcomes in older people. The primary focus of the review was to identify instruments suitable for application with older people within economic evaluations conducted in the aged care sector.MethodsOnline databases searched were PubMed, Medline, Scopus, and Web of Science, PsycInfo, CINAHL, Embase and Informit. Studies that met the following criteria were included: 1) study population exclusively above 65 years of age 2) measured health status, health related quality of life or quality of life outcomes more broadly through use of an instrument developed for this purpose, 3) used a generic preference based instrument or an older person specific preference based or non-preference based instrument or both, and 4) published in journals in the English language after 2000.ResultsThe most commonly applied generic preference based instrument in both the community and residential aged care context was the EuroQol - 5 Dimensions (EQ-5D), followed by the Adult Social Care Outcomes Toolkit (ASCOT) and the Health Utilities Index (HUI2/3). The most widely applied older person specific instrument was the ICEpop CAPability measure for Older people (ICECAP-O) in both community and residential aged care.ConclusionIn the absence of an ideal instrument for incorporating into economic evaluations in the aged care sector, this review recommends the use of a generic preference based measure of health related quality of life such as the EQ-5D to obtain quality adjusted life years, in combination with an instrument that has a broader quality of life focus like the ASCOT, which was designed specifically for evaluating interventions in social care or the ICECAP-O, a capability measure for older people.Electronic supplementary materialThe online version of this article (doi:10.1186/s12955-015-0372-8) contains supplementary material, which is available to authorized users.

Brand versus generic medications: A disease state approach to identify patients’ perceptions and concerns

Background Medication cost does not solely dictate patient preference of brand name or generic medications. However, few studies have been completed to assess the influence of other factors, such as demographics and chronic diseases. Methods Responses were collected from 347 English speaking adults in the United States, age 18 and older, pertaining to demographics, diagnosed conditions, preference and knowledge of brand and generic medications. Results Participants with COPD, epilepsy, and high cholesterol felt that their condition influenced their preference of brand or generic medication more as compared with participants diagnosed with other conditions. If cost was not an issue, then 46% of respondents would prefer the brand name product, while 18% of participants felt their preference depended on the condition that the medication was intended to treat. Roughly one-third of the participants did not fully understand why there are both brand name and generic medications, and education level did not impact comprehension significantly. Conclusion The incentive for generic preference over brand name is primarily due to cost differences. However, improvement of education pertaining to brand/generic equivalency is needed, as some patients still believe brand name medications are superior to generic alternatives for certain disease states.

What future for the French retail generic market? Can generic companies survive?

The worldwide generic market doubled, in volume, over the past 10 years. It appears, at first sight, very attractive. However, generic companies’ profitability is on average lower than that of R&amp;D-based companies and their future risk of deterioration higher due to stronger commoditisation. The attractiveness of generic markets being significantly different from one country to the other, thus, the author proposes an approach, that he has applied to the French market, to assess country-specific business opportunities and threats. The analysis of different stakeholders shows that the French generic market development has been driven by a series of governmental measures to encourage physicians to prescribe, pharmacists to substitute and deliver, and patients to accept generic products. Physicians, who did not play a major role so far, could be asked to prescribe significantly more generic products, or at least more genericized brands, so that pharmacists can substitute them. Retail pharmacists, who substitute more than 70% of genericized brands, have been instrumental in the development of generic products and will be asked to further increase their substitution rate. Patients, of whom almost two-third accepts generic products, are not expected to change their attitude over the short term. Generic companies, operating in France are in danger due to anticipated price cuts of the government in one hand, and the refusal of retail pharmacists to accept any decrease of their discounts, in the other hand. In such an increasingly unattractive environment, survivors will be the biggest generic companies which will manage to secure the lowest production costs and the highest Generic Preference Mix index.

Mindfulness for the self-management of fatigue, anxiety, and depression in women with metastatic breast cancer: a mixed methods feasibility study.

The impact of living with metastatic breast cancer (MBC) is considerable and psychosocial support can be beneficial. Mindfulness-based stress reduction (MBSR) can help self-management of anxiety, depression, quality of life (QoL), and fatigue and has been evaluated in early-stage breast cancer but not MBC. This study investigated the acceptability and feasibility of providing MBSR for women with MBC and of introducing MBSR into a National Health Service (NHS) setting. A mixed methods convergent design was used. Eligible women with MBC, an Eastern Cooperative Oncology Group (ECOG) score of 0 to 2, stable disease, and life expectancy of at least 6 months were invited to attend (by their oncologist) an 8-week MBSR course. Qualitative interviews with patients, a focus group, and interview with NHS staff were held to explore acceptability and feasibility of MBSR. Questionnaires at baseline, during (weeks 4, 8), and after (weeks 16, 24) the course measured fatigue, anxiety and depression, mindfulness, disease-specific QoL, and generic preference based QoL. Of 100 women approached, 20 joined the study. One woman dropped out prior to the intervention due to illness progression. Nineteen women took part in 3 MBSR courses. Recruitment to 2 of the 3 courses was slow. Commitment to 8 weeks was a reason for non-participation, and proved challenging to participants during the course. Participants found the course acceptable and reported many cumulative and ongoing benefits. These included feeling less reactive to emotional distress and more accepting of the disruption to life that occurs with living with MBC. There was high attendance, completion of course sessions, adherence to home practice, excellent follow-up rates, and high questionnaire return rates. MBSR was acceptable to MBC patients, who perceived benefits such as improved anxiety and QoL; but the MBSR course requires a considerable time commitment. There is scope to tailor the intervention so that it is less intensive.

MA3 Mapping the Diabetes Health profile (DHP-18) onto the EQ-5D and SF-6D Generic Preference Based Measures of Health

MA3 Mapping the Diabetes Health profile (DHP-18) onto the EQ-5D and SF-6D Generic Preference Based Measures of Health

Organic food choices and Protection Motivation Theory: Addressing the psychological sources of heterogeneity

Addressing generic preference heterogeneity for foods has been shown to improve statistical model fit, but it is relatively poor in terms of insights as to the sources of heterogeneity. The study presents various results from latent class analysis of stated choice data of food choice using carrots as a prototype. We evaluate the coherence of the underlying preference structure using as criteria the psychological constructs used in Protection Motivation Theory (PMT) ( Rogers, 1975) elicted by means of attitudinal responses. We estimate latent classes based on such responses using both simultaneous and two-stage procedures. The latter performs a segmentation more in line with expectations derived from PMT, in terms of both the number of classes and their preference structures. While simultaneous estimation improves the statistical fit it yields class structures that appear less reasonable.

Health outcome and economic measurement in breast cancer surgery: challenges and opportunities

As breast cancer surgery techniques continue to advance, treatment options continue to increase, bringing with them increased scrutiny of health outcomes and healthcare cost. In addition, patients are becoming more involved in their own medical care and are demanding meaningful data to help them better understand expected outcomes. With these changes and advancements, there is a growing emphasis on evidence-based practice. In this article, we focus on scientific considerations, challenges to and opportunities for improving outcome measurement related to breast cancer surgery. There are two main messages from this article. First, until recently, rigorously developed specific patient-reported outcome (PRO) measures for breast cancer surgery patients have not been available for use. However, with the recent introduction of new PRO measures, such as the BREAST-Q, there is now good potential to collect useful outcome data on patient satisfaction and health-related quality of life, and to better understand the relative impact of different surgical procedures, decision making and clinical practice on patient outcome. Thus, PRO research using rigorously developed breast cancer surgery-specific measures is in its infancy, but growing steadily. Second, there is a great need but lack of specific health economic measures developed for use in breast cancer surgery research. In fact, research into the economic evaluation of breast cancer surgery is an area that has received less attention than that of PRO measure development, but there is good opportunity to expand this area of research in breast cancer surgery. Further studies are required to gain a clearer view of the role that generic preference and utility measures could play, how best to synthesize health-related quality of life and economic metrics data, and the potential use of new disease-specific tools.

The health related quality of life of people living with HIV/AIDS in sub-Saharan Africa - a literature review and focus group study

BackgroundWhile health outcomes of HIV/AIDS treatments in terms of increased longevity has been the subject of much research, there appears to be very limited research on the improved health related quality of life (HRQL) that can be applied in cost-utility analyses in Africa south of the Sahara (SSA). Most of the literature that does exist present HRQL measured by disease specific instruments, but such data is of little use as input to economic evaluations.MethodsA systematic review of the literature on HRQL weights for people living with HIV/AIDS in Africa was performed, and the findings are presented and interpreted. We also use focus group discussions in panels of clinical AIDS experts to test the preference based on a generic descriptive system EQ-5D. We contrast quality of life with and without antiretroviral treatment (ART), and with and without treatment failure.ResultsIn only four papers were the HRQL weights for HIV/AIDS in sub-Saharan Africa estimated with generic preference based methodologies that can be directly applied in economic evaluation. A total of eight studies were based on generic health profiles. While such 'health profiles' are not preference based, the scores could potentially be transformed into health state utilities. Most of the available literature (20 papers) utilized disease specific instrument, which are not applicable for economic evaluation.The focus group discussions revealed that HRQL weights are strongly correlated to disease stage. Furthermore, clinical experts consistently report that ART has a strong positive impact on the HRQL of patients, although this effect appears to rebound in cases of drug resistance.ConclusionsEQ-5D appears to be an appropriate tool for measuring and valuing HRQL of HIV/AIDS in Africa. More empirical research is needed on various methodological aspects in order to obtain valid and reliable HRQL weights in economic evaluations of HIV/AIDS prevention and treatment interventions.

Generic Preferences over Subsets of Structured Objects

Various tasks in decision making and decision support systems require selecting a preferred subset of a given set of items. Here we focus on problems where the individual items are described using a set of characterizing attributes, and a generic preference specification is required, that is, a specification that can work with an arbitrary set of items. For example, preferences over the content of an online newspaper should have this form: At each viewing, the newspaper contains a subset of the set of articles currently available. Our preference specification over this subset should be provided offline, but we should be able to use it to select a subset of any currently available set of articles, e.g., based on their tags. We present a general approach for lifting formalisms for specifying preferences over objects with multiple attributes into ones that specify preferences over subsets of such objects. We also show how we can compute an optimal subset given such a specification in a relatively efficient manner. We provide an empirical evaluation of the approach as well as some worst-case complexity results.

Putting the ‘Q’ in depression QALYs: a comparison of utility measurement using EQ-5D and SF-6D health related quality of life measures

Generic health preference measures that capture quality of life improvements in depression are important for economic analysis of new technologies. To compare two widely used preference measures, the EQ-5D and SF-6D, in terms of their dimensions and health utility values. EQ-5D and SF-6D data collected from 114 patients with depression, who participated in a cluster, randomised controlled trial to evaluate a collaborative care intervention in UK, primary care practices. Utilities were examined across the whole sample and by level of depression severity using the PHQ-9. Depression was associated with disutility at baseline. At 3 month follow-up mean utility increased 0.147 for EQ-5D and 0.082 for SF-6D. Health gains were observed in patients in remission from depression and those with the mildest level of depression severity. Both generic preference measures were sensitive to health gains in depression within a relatively short follow-up period; larger health gain was observed for the EQ-5D.

Habitat use and preferences of cetaceans along the continental slope and the adjacent pelagic waters in the western Ligurian Sea

The physical habitat of cetaceans occurring along the continental slope in the western Ligurian Sea was investigated. Data were collected from two different sighting platforms, one of the two being a whale-watching boat. Surveys, conducted from May to October and from 1996 to 2000, covered an area of approximately 3000 km 2 with a mean effort of about 10,000 km year −1. A total of 814 sightings was reported, including all the species occurring in the area: Stenella coeruleoalba, Balaenoptera physalus, Physeter macrocephalus, Globicephala melas, Grampus griseus, Ziphius cavirostris, Tursiops truncatus, Delphinus delphis. A Geographic Information System was used to integrate sighting data to a set of environmental characteristics, which included bottom gradient, area between different isobaths, and length and linearity of the isobaths within a cell unit. Habitat use was analysed by means of a multi-dimensional scaling, MDS, analysis. Significant differences were found in the habitat preference of most of the species regularly occurring in the area. Bottlenose dolphin, Risso's dolphin, sperm whale and Cuvier's beaked whale were found strongly associated to well-defined depth and slope gradient characteristics of the shelf-edge and the upper and lower slope. The hypothesis of habitat segregation was considered for Risso's dolphin, sperm whale and Cuvier's beaked whale. Canonical discriminant functions using depth and slope as predictors outlined clear and not overlapping habitat preferences for Risso's dolphin and Cuvier's beaked whale, whereas a partial overlapping of the habitat of the other two species was observed for sperm whale. Such a partitioning of the upper and lower slope area may be the result of the common feeding habits and suggests a possible competition of these three species. A temporal segregation in the use of the slope area was also observed for sperm whales and Risso's dolphins. Fin whales, and the occasionally encountered common dolphin and long-finned pilot whales showed a generic preference for the pelagic area. In contrast, striped dolphins were generically found in any type of physical habitat, in agreement with their known opportunistic feeding habits. Interannual and seasonal movements were observed in fin whales and in all the other regularly occurring species, indicating that changes in the environmental and climatological conditions have a remarkable influence on habitat use. Finally, a relationship between group size and habitat use was observed for the species known to forage in groups. This relationship suggests that these species respond to environmental variability either by moving within the study area or by behavioural changes that may reflect changes in the preference for prey items.