Get 20M+ Full-Text Papers For Less Than $1.50/day. Subscribe now for You or Your Team.

Learn More →

Genetic nurse counsellors can be an acceptable and cost-effective alternative to clinical geneticists for breast cancer risk genetic counselling. Evidence from two parallel randomised controlled equivalence trials

Genetic nurse counsellors can be an acceptable and cost-effective alternative to clinical... British Journal of Cancer (2006) 95, 435 – 444 & 2006 Cancer Research UK All rights reserved 0007 – 0920/06 $30.00 www.bjcancer.com Genetic nurse counsellors can be an acceptable and cost-effective alternative to clinical geneticists for breast cancer risk genetic counselling. Evidence from two parallel randomised controlled equivalence trials ,1 1 2 3 4 4 5 5 N Torrance , J Mollison , S Wordsworth , J Gray , Z Miedzybrodzka , N Haites , A Grant , M Campbell , 1 3 6 MS Watson , A Clarke and B Wilson 1 2 Department of Public Health, Medical School, University of Aberdeen, Foresterhill, Aberdeen AB25 2ZD, UK; Health Economics Research Centre, 3 4 University of Oxford, Oxford OX3 7LF, UK; Institute of Medical Genetics, University Hospital of Wales, Cardiff CF14 4XN, UK; Department of Medicine & Therapeutics, University of Aberdeen, Aberdeen AB25 2ZD, UK; Health Services Research Unit, University of Aberdeen, Aberdeen AB25 2ZD, UK; Department of Epidemiology & Community Medicine, University of Ottawa, 451 Smyth Road, Ottawa, Ontario, Canada K1H 8M5 This study compared genetic nurse counsellors with standard services for breast cancer genetic risk counselling services in two regional genetics centres, in Grampian region, North East Scotland and in Cardiff, Wales. Women referred for genetic counselling were randomised to an initial genetic counselling appointment with either a genetic nurse counsellor (intervention) or a clinical geneticist (current service, control). Participants completed postal questionnaires before, immediately after the counselling episode and 6 months later to assess anxiety, general health status, perceived risk and satisfaction. A parallel economic evaluation explored factors influencing cost-effectiveness. The two concurrent randomised controlled equivalence trials were conducted and analysed separately. In the Grampian trial, 289 patients (193 intervention, 96 control) and in the Wales trial 297 patients (197 intervention and 100 control) returned a baseline questionnaire and attended their appointment. Analysis suggested at least likely equivalence in anxiety (the primary outcome) between the two arms of the trials. The cost per counselling episode was d11.54 less for nurse-based care in the Grampian trial and d12.50 more for nurse-based care in Cardiff. The costs were sensitive to the grade of doctor (notionally) replaced and the extent of consultant supervision required by the nurse. In conclusion, care based on genetic nurse counsellors was not significantly different from conventional cancer genetic services in both trial locations. British Journal of Cancer (2006) 95, 435–444. doi:10.1038/sj.bjc.6603248 www.bjcancer.com Published online 11 July 2006 & 2006 Cancer Research UK Keywords: genetic counselling; familial cancer; breast cancer; randomised control trial; equivalence; anxiety Familial cancer is an area in which genetic knowledge has to increased lifetime risks of developing the disease, often at an progressed rapidly over the past 10 years, and where patient earlier age than sporadic disease (Claus et al, 1994). However, demand for genetics services for information, counselling and these mutations are implicated in around only 5–10% of breast mutation testing has increased dramatically. Even with the cancer cases, with the result that genetic testing is not appropriate introduction of referral guidelines, regional genetics clinics are for the majority of women with a family history of breast cancer. facing the challenge of increasing demand within fixed resources Genetic counselling services aim to identify individuals who have a (Kinmonth et al, 1998), and these difficulties may be exacerbated significantly increased genetic risk of cancer and counsel them as new mutations are identified and further demands are made on about appropriate risk management to reduce morbidity and genetics clinics. mortality (Fry et al, 2003). Access to surveillance, testing and other Breast cancer is the most common cancer affecting UK women, interventions is usually dependent on the outcome of the initial resulting in a lifetime risk of one in nine of developing the disease genetic counselling and risk assessment. (Office for National Statistics, 2005). The discovery of breast Recent research in the area of breast cancer genetic risk cancer susceptibility mutations has attracted widespread publicity counselling has generally concentrated on women’s psychological and women with a family history of breast cancer dominate status and risk perception. Evidence from systematic reviews referrals throughout the UK for any cancer genetic risk counselling suggests that genetic counselling does not appear to have any (Wonderling et al, 2001; Hopwood et al, 2004). Mutations in the adverse effect on psychological outcomes (Meiser and Halliday, known inherited susceptibility genes, BRCA1 and BRCA2, give rise 2002; Butow et al, 2003; Braithwaite et al, 2004), and may convey some short-term benefits in decreasing general anxiety and in improving the accuracy of women’s perceptions of their personal *Correspondence: Dr N Torrance; E-mail: n.torrance@abdn.ac.uk risk. Historically, clinical genetic risk assessments in the UK have Received 31 March 2006; revised 5 June 2006; accepted 8 June 2006; published online 11 July 2006 been conducted largely by medical doctors, with a variable amount Clinical Studies Clinical Studies Genetic nurse counsellors can be an acceptable and cost-effective N Torrance et al of support for genetic counselling provided by genetic nurses and Eligible patients were randomised 2 : 1 to intervention: control counsellors, social workers and psychologists as well. Many groups. This uneven allocation was chosen to give greater specialist centres have recently begun to expand their genetic risk experience of the novel intervention and was taken account of in counselling capacity by recruiting genetic nurses or associates the sample size calculation. Postal questionnaires were adminis- (Skirton et al, 1998). This trend is likely to continue. However, tered to participants at baseline (before the clinic appointment), there is a lack of formal evaluation of the effectiveness of this immediately following the counselling episode (FU1), and 6 strategy and the lack of well designed and conducted randomised months later (FU2). controlled trials of genetic service models has been noted in a number of recent publications (Butow et al, 2003; Braithwaite et al, Standard care in two trial locations (clinical geneticist) 2004; Hopwood et al, 2004; NICE, 2004). We therefore conducted a pragmatic evaluation comparing As indicated above, the control arms reflected current standard of genetic nurse counsellors as an alternative to a physician-based care in each location, and no attempt was made to alter this. The service (standard care) in assessing patients newly referred for process for patients referred for genetic assessment was similar in genetic counselling for risk of breast cancer. We focused on both centres: family history taking, pedigree construction, patient-centred outcomes, and acceptability to referring general confirmation of cancer diagnoses in affected relatives, risk practitioners, and conducted a concurrent economic evaluation. In assessment, genetic counselling with advice on preventive order to increase the generalisability of our findings, we conducted strategies and, if appropriate screening, prophylactic surgery, the evaluation as two randomised controlled equivalence trials and/or genetic testing. Initially, the clinical geneticist took the with differing patient population and service characteristics. family history, constructed the pedigree, confirmed cancer diagnoses, etc in both centres. In Grampian, clerical and nursing support was available to enable the family tree to be constructed in advance of the clinic appointment. In Wales the clinical geneticist MATERIALS AND METHODS did all of the preclinic preparation. In both settings, clinic appointments involved a clinical geneticist and lasted approxi- Participants and procedures mately 45 min. Two concurrent randomised controlled trials were conducted to assess the equivalence of care provided by a genetic nurse counsellor with clinical geneticist-based (standard) care. The Intervention (genetic nurse counsellors) study locations were the Grampian University Hospitals NHS In both trials, the nurse counsellor made the initial contact with Trust Clinical Genetics Service (in Aberdeen) and the Wales the patient. Each followed the format of the standard care Genetics Service (in Cardiff). The study was designed with a described above for their own centre, with a nurse counsellor pragmatic approach; that is, the goal was to gain an insight into the substituting for clinical geneticist input. Each clinic was therefore overall effects of substituting a nurse counsellor-based approach free to make its own organisational arrangements for the nurse-led for a physician-based approach, while accepting that contextual counselling arm, so long as there were adequate minimum factors inevitably vary between health care settings. Thus, there supervision arrangements, a clinical geneticist was professionally were differences in regular (doctor-based) care in the two trial and managerially accountable for the patient care delivered by a locations, but both represented an acceptable ‘standard of care’ for nurse counsellor, and the same risk assessment algorithms were breast cancer genetic risk assessment and counselling within the followed within each clinic (by both nurse counsellors and clinical NHS. Similarly, the intervention arms reflected nurse counsellor- geneticists). The focus of the study was the ‘counselling episode’: based care developed and managed according to local protocols the time from appointment scheduling until the patient was either and supervisory arrangements. As a clinical geneticist was formally discharged or further management arrangements were made. This responsible in each location for patient care, whether it was provided the ‘boundary’ to the intervention – all activities from delivered by a doctor or a nurse counsellor, we were not concerned this point on were part of the ‘usual’ cancer genetics service. with assessing performance of the nurse counsellors in relation to Participants who were considered likely candidates for genetic technical procedures such as pedigree construction or calculation testing (i.e. high risk) were referred for further follow-up with a of risk estimates. Rather, we concentrated on other outcomes of consultant geneticist. In both trials, a consultant geneticist actively interest, including provider and patient satisfaction. By imple- supervised the nurse counsellors, meeting with them on a weekly menting trial procedures rigorously, collecting identical data sets, basis. Training was delivered according to the needs of the nurse but analysing the two trials separately, we were able to promote counsellors when they were appointed. internal validity within each trial but also comment on external validity (i.e. generalisability) (Pocock, 1983). Patients were eligible if they were newly referred because of Data collection concern about family history of breast cancer, were aged 18 years or over, and literate in English. Patients previously affected by Sociodemographic data Sociodemographic data were collected breast cancer were included. Patients were excluded if they had from the women in the baseline questionnaire, and included age, previously attended the genetics clinic or were known to be a marital status, number of children, ethnicity and educational level. member of a family in which a BRCA1 or BRCA2 mutation had been previously identified. Psychological outcomes The primary outcome was patient Eligible patients were identified from referral letters and consent anxiety, measured using the six-item short-form of the state scale to contact was obtained from each patient’s GP. Before their clinic of the Spielberger State-Trait Anxiety Inventory (STAI; Marteau appointment, patients were sent a letter inviting their participa- and Bekker, 1992). The short version produces result comparable tion, an information sheet explaining the study and a consent with the full state scale and has been used in other genetic form. Those who consented were randomised to the intervention counselling research (Miedzybrodzka et al, 1995; Bish et al, 2002). (genetic nurse counsellor-based care) or control group (clinical We also included the Hospital Anxiety and Depression Scale geneticist-based care). The random allocation schedule sequence (HADS; Zigmond and Snaith, 1983), which gives separate measures for each trial was computer-generated and concealed within a of anxiety and depression assessed over the past week (range of Microsoft Access database. To avoid possible contamination, scores from 0 to 21, seven items for anxiety and seven for members of the same family were allocated to the same trial group. depression). British Journal of Cancer (2006) 95(4), 435 – 444 & 2006 Cancer Research UK Genetic nurse counsellors can be an acceptable and cost-effective N Torrance et al Health-related quality of life The Short Form 36 (SF-36) health but before data were released for analysis. For anxiety as measured survey instrument was used to measure perception of general by HADS, and the role-emotional and mental health domains of health status (Ware et al, 1994). It measures patients’ rating of the SF36, we defined the equivalence limit as one-third of a their own health status in eight areas or domains: physical standard deviation of the baseline scores. functioning, role physical, role emotional, social functioning, We analysed data using the ‘intention to treat’ principle, in mental health, vitality, bodily pain and general health perception. order to minimise the effect of selection bias and tested the For each dimension item scores range from 0 (worst possible robustness of the findings by comparing the findings to a health state) to 100 (best possible health state). ‘treatment-received’ analysis. Analysis was performed using SPSS v11.0. Perceived risk of breast cancer We assessed participants’ perceived risk of breast cancer at all three time points with an Sample size The primary outcome was anxiety as assessed by item used in previous research (Lerman et al, 1993; Lloyd et al, STAI (short form), with an equivalence limit of 74 units. We 1996; Watson et al, 1999; Brain et al, 2000). Women were asked to required 214 participants in the intervention group and 107 in the assess their own risk of developing breast cancer relative to a control group to allow detection of equivalence (two sided), with notional ‘average’ woman, on a five-point scale (‘much lower/ 80% power at the 5% significance level. This assumed a standard lower/about the same/higher/much higher’). Women previously deviation of 12 units for STAI (Marteau et al, 1990), and affected by breast cancer were asked to skip this section of the incorporated the 2 : 1 allocation ratio. questionnaire to avoid potential confusion. Economic evaluation Knowledge of breast cancer risk factors We assessed participants’ understanding of risk factors for breast cancer at all three time A cost analysis was conducted alongside the trials, adopting a points. Women were asked how strongly they agreed with three societal perspective to include health service and patient costs. The statements on specific causes of breast cancer (stress, having a ‘cost per patient counselling episode’ for each woman in both trials first-degree relative with breast cancer and minor injury). We used was calculated. Data on health service resource use (staff, a five-point scale (strongly disagree/disagree/not sure/agree/ consumables, rooms and equipment) were collected and local unit strongly agree). For each of these three potential risk factors, costs applied where available, otherwise national rates were used. agreement would indicate a response inconsistent with the Staff time included outpatient appointments, preparation, meet- education content offered during the genetic counselling appoint- ings and time in the nurse counsellor arm for consultant ment. geneticists discussing patient cases with the nurses. The mid- point of salary scales was used, adding employers’ on-costs at 13%. Patient satisfaction and acceptability to referring GPs Patient An equivalent annual cost was calculated for equipment items such satisfaction was measured using a modified version of the as computers, over relevant life-spans for the items, using a 6% Satisfaction with Genetic Counselling Questionnaire developed discount rate. A total cost per counselling episode for each woman by Shiloh et al (1990) and was incorporated into both follow-up was calculated by multiplying the unit cost per single appointment questionnaires. The scale assesses three dimensions of patient by the number of appointments in each trial arm. Sensitivity satisfaction: (1) instrumental (satisfaction with the doctor/nurse’s analysis tested the robustness of the findings and data on patient competence), (2) affective (satisfaction with the doctor/nurse’s time and travel costs were also collected. All costs are reported in personal qualities) and (3) procedural (satisfaction with adminis- Sterling (d) for the price year 2006. trative procedures, such as waiting time and staff conduct). The Approval for the study was obtained from the Joint Ethics acceptability of the genetic nurse counsellors to referring GPs was Committee of Aberdeen University and Grampian Health Board, assessed towards the end of the study. A short questionnaire was and the research ethics committees of Bro Taf and Iechyd used to ascertain whether the referring GPs had noticed any Morgannwg health authorities. difference (including deterioration) in the service provided by the genetics department for their patients, if they would be happy for their patients to be seen by a genetic nurse counsellor in the future RESULTS and an overall rating of their satisfaction with the genetics services. Participants Statistical analysis Grampian In total, 517 referred patients were considered for recruitment (Figure 1). Of the 342 (66%) patients who consented Descriptive statistics were used to describe the study participants. and were randomised (227 intervention, 115 control), 289 (84%) As this was an equivalence trial, a confidence interval (CI) returned a baseline questionnaire and attended clinic (193 (85%) approach was used (Jones et al, 1996). The difference in outcomes intervention, 96 (83%) control). In total, 17 did not receive the was calculated, adjusting for differences in baseline scores using allocated management because of administrative errors (6), joint multiple linear regression. For each outcome, the 95% CIs around family appointments (4) or decision by the head of service (7). the difference were calculated. An outcome was considered ‘equivalent’ when the 95% CI for the difference between the intervention and control group fell completely within a prede- Wales In total, 464 patients were considered for recruitment termined equivalence limit (Jones et al, 1996). Where the 95% CI (Figure 1). Of the 373 (80%) patients who consented and were for the observed difference fell completely outside the equivalence randomised (247 intervention, 126 control), 297 (80%) returned a limit, we considered the outcomes would be ‘nonequivalent’. baseline questionnaire and attended clinic (197 (80%) interven- Where the 95% CI for the observed difference overlapped the tion, 100 (79%) control). Six women did not receive the allocated equivalence limit, the result would be uncertain. For the primary management because of administrative errors (2) and joint family outcome, the STAI score, we defined an a priori strict limit of appointment (4). ‘equivalence’ of 74 units (used in the sample size calculation Baseline data are presented for women who returned a baseline below), and a ‘likely equivalence’ limit of 710 units (which the questionnaire and attended a clinic appointment (Table 1). The trial steering group considered more reflective of actual clinical randomised groups were generally similar in terms of demo- practice). The strict limit was defined before the study started, and graphic characteristics. Within each trial, similar proportions of the likely equivalence limit was determined after the study began, participants allocated to each arm perceived themselves at elevated & 2006 Cancer Research UK British Journal of Cancer (2006) 95(4), 435 – 444 Clinical Studies Genetic nurse counsellors can be an acceptable and cost-effective N Torrance et al British Journal of Cancer (2006) 95(4), 435 – 444 & 2006 Cancer Research UK Progress of participants through the trials Grampian Wales Patients assessed for eligibility (n =464) Patients assessed for eligibility (n =517) Not randomised (n =91) Not randomised (n =175) Reasons: refused to take part (n =26); did not fulfil inclusion Reasons: refused to take part (n =48); did not fulfil inclusion criteria (n =6), no response to recruitment (n =59) criteria (n =28); no response to recruitment (n =99) Randomised (n =373) Randomised (n =342) Allocated to doctor (control) arm (n = Allocated to genetic nurse (intervention) Allocated to genetic nurse (intervention) Allocated to doctor (control) arm (n =115) 126) arm (n =247) arm (n =227) Returned baseline questionnaire and Returned baseline questionnaire and Returned baseline questionnaire and Returned baseline questionnaire and attended clinic (n =96) attended clinic (n =197) attended clinic (n =100) attended clinic (n=193) Received allocated intervention (n =92) Received allocated intervention (n=181) Received allocated intervention (n =191) Received allocated intervention (n =100) Immediate follow-up (n =175) Immediate follow-up (n =83) Immediate follow-up (n =169) Immediate follow-up (n =85) Lost to follow-up (n =13) Lost to follow-up (n =18) Lost to follow-up (n =27) Lost to follow-up (n =14) Withdrew (n =1) Withdrew (n =1) Six-month follow-up (n =163) Six-month follow-up (n =74) Six-month follow-up (n =150) Six-month follow-up (n =73) Lost to follow-up (n =12) Lost to follow-up (n =9) Lost to follow-up (n =19) Lost to follow-up (n =12) Completed trial (n =73) Completed trial (n =150) Completed trial (n =74) Completed trial (n =163) Notes : Notes : a e Did not return baseline questionnaire (n =20), did not attend clinic (n =11), moved away (n =1), Did not return baseline questionnaire (n =39), did not attend clinic (n =10), administrative error (n =1), subsequently found to be not eligible (n =2). subsequently found to be not eligible (n =2). b f Attended appointment with Doctor arm (n =12). Attended appointment with Doctor arm (n =6). c g Did not return baseline questionnaire (n =15), did not attend clinic (n =4). Did not return baseline questionnaire (n =22), did not attendclinic (n =2), subsequently found to be not eligible (n =2). Attended appointment with nurse arm (n =4). Figure 1 Progress of participants through the trials. Clinical Studies Genetic nurse counsellors can be an acceptable and cost-effective N Torrance et al Table 1 Characteristics of study participants at baseline risk, although Wales participants were generally more likely to view themselves as at elevated risk than Grampian participants. Grampian Wales The actual estimated lifetime risk of breast cancer for participants only became available after each initial individual counselling Nurse Nurse episode was complete. In the Grampian trial, participants in the counsellor Geneticist counsellor Geneticist intervention arm were more likely than those in the control arm to be assessed as being at elevated risk, whereas the converse was No. of patients 193 96 197 100 observed for the Wales trial. The analysis by treatment received Age (years): mean 40.7 (10.3) 41.4 (9.4) 39.8 (10.2) 39.0 (9.3) produced only minor differences in results compared with the (s.d.) Married/cohabiting: n 151 (80.3) 74 (77.9) 161 (81.7) 84 (84.8) analysis by intention-to-treat, therefore the data are not reported (%) in this paper. With children: n (%) 157 (83.5) 73 (76.0) 157 (79.7) 73 (73.0) Post-secondary 76 (39.4) 43 (44.8) 73 (37.1) 42 (42.0) education: n (%) Anxiety and general health status Referral source: n (%) Table 2 summarises the primary outcomes for the two trials. There General 133 (68.9) 63 (65.6) 117 (59.4) 56 (56.0) were small but consistent baseline differences between the practitioner Grampian and Wales study populations, but generally scores were Breast surgeon 35 (18.1) 19 (19.8) 73 (37.1) 42 (42.0) comparable between intervention and control arms within each Breast screening 17 (8.8) 10 (10.4) — — trial. With respect to STAI, all adjusted point estimates for the clinic differences between intervention and control groups met the Other 8 (4.1) 4 (4.2) 7 (3.5) 2 (2.0) definition of, at least, ‘likely equivalence’. a a b b For anxiety and depression as measured by HADS, a priori n (%) perceiving 130/181 65/91 (71) 119/ 179 76/90 (84) equivalence limits of one-third of the baseline standard deviation themselves to be at (72) (83) elevated risk at were calculated from the data as 71.4 (Grampian) and 71.5 baseline (Wales) for anxiety and 71.2 (both trials) for depression. These c c d d n (%) assessed at 173/192 74/95 (78) 145/197 84/100 (84) are close to the smallest possible difference in score for an clinic at elevated risk (90) (74) individual, which is 71 point. The results of all analyses were consistent with ‘equivalence’. Also, shown are the observed For perceived risk, participants previously affected by breast cancer are excluded: a b differences in the SF36 role-emotional and mental health domains. Seven nurse counsellor, one geneticist. 10 nurse counsellor, three geneticist. For The equivalence limits were calculated as 711.4 and 713.1 for the assessed risk, participants previously affected by breast cancer are included: Five nurse counsellor, one geneticist. 10 nurse counsellor, three geneticist. role-emotional score, and 76.0 and 76.3 for the mental health Table 2 Psychological outcomes, mean (s.d.) Grampian Wales b b Nurse Equivalence Difference Nurse Equivalence Difference Outcomes counsellor Geneticist limit (95% CI) counsellor Geneticist limit (95% CI) STAI Baseline 37.3 (13.6) 36.5 (12.8) 40.9 (15.1) 40.0 (14.5) Immediately after episode 36.4 (14.0) 34.4 (14.0) 74.0 0.8 (2.1 to 3.7) 38.1 (14.9) 38.9 (15.6) 74.0 1.5 (4.5 to 1.5) Six months after episode 36.0 (13.5) 32.1 (11.7) 2.9 (0.2 to 5.9) 38.9 (14.9) 38.1 (14.1) 0.6 (2.9 to 4.1) HADS Anxiety Baseline 6.7 (4.3) 6.4 (4.5) 8.1 (4.7) 7.4 (4.2) Immediately after episode 6.3 (4.3) 5.5 (3.9) 71.4 0.5 (0.4 to 1.3) 7.0 (4.9) 7.1 (4.8) 71.5 0.4 (1.3 to 0.5) Six months after episode 6.2 (4.4) 5.5 (3.7) 0.1 (0.7 to1.0) 7.4 (4.7) 6.4 (4.1) 0.5 (0.6 to 1.5) HADS Depression Baseline 3.9 (3.7) 3.4 (3.4) 4.5 (3.7) 4.2 (3.8) Immediately after episode 3.5 (3.6) 2.9 (2.8) 71.2 0.3 (0.4 to 1.0) 4.0 (3.8) 4.0 (3.8) 71.2 0.2 (1.0 to 0.5) Six months after episode 3.4 (3.6) 2.8 (2.9) 0.3 (0.5 to 1.0) 4.5 (4.1) 3.9 (3.8) 0.6 (0.4 to 1.5) SF36 Role emotional Baseline 80.5 (34.6) 82.6 (33.4) 74.4 (38.7) 71.0 (40.6) Immediately after episode 81.6 (35.2) 82.5 (33.2) 711.4 1.9 (6.3 to 10.1) 74.8 (39.5) 71.5 (40.0) 713.1 2.9 (6.9 to 12.7) Six months after episode 80.3 (35.9) 86.0 (30.7) 2.5 (11 to 5.9) 74.9 (38.7) 73.1 (42.2) 0.5 (9.4 to 10.5) SF36 Mental health Baseline 71.0 (18.2) 73.6 (17.7) 67.3 (18.8) 68.4 (19.3) Immediately after episode 72.2 (18.6) 74.4 (17.7) 76.0 0.6 (2.9 to 4.1) 68.8 (20.5) 68.0 (21.3) 76.3 1.3 (2.7 to 5.2) Six months after episode 72.3 (18.4) 77.4 (14.9) 2.7 (6.5 to 1.2) 67.1 (21.1) 67.4 (21.1) 0.3 (4.2 to 4.8) a b c d Set at 74.0 for STAI, 1/3 of baseline s.d.s otherwise. Adjusted for baseline. Higher score indicates greater level of anxiety. Higher score indicates greater level of depression. Higher score indicates better health state. & 2006 Cancer Research UK British Journal of Cancer (2006) 95(4), 435 – 444 Clinical Studies Clinical Studies Genetic nurse counsellors can be an acceptable and cost-effective N Torrance et al Table 3 Health related quality of life, SF-36 scores : mean (s.d.) Grampian Wales b b Time Nurse counsellor Geneticist Difference (95% CI) Nurse counsellor Geneticist Difference (95% CI) Physical functioning Baseline 88.9 (18.6) 85.4 (21.8) 83.5 (21.8) 88.9 (18.3) FU1 88.2 (19.1) 88.6 (18.7) 0.5 (3.3 to 2.2) 84.9 (20.9) 88.0 (22.5) 0.3 (3.7 to 4.3) FU2 87.9 (18.5) 86.4 (21.3) 0.2 (2.3 to 2.7) 84.6 (21.3) 88.6 (20.1) 0.5 (5.5 to 4.5) Social functioning Baseline 84.0 (23.4) 84.6 (22.1) 77.7 (24.4) 79.8 (26.2) FU1 83.4 (23.9) 85.2 (21.2) 0.6 (5.7 to 4.5) 78.9 (24.8) 79.9 (25.4) 0.3 (4.9 to 5.4) FU2 84.7 (21.8) 87.2 (23.2) 1.0 (6.4 to 4.4) 78.0 (26.9) 80.1 (26.4) 1.0 (7.6 to 5.5) Role physical Baseline 87.1 (29.0) 86.1 (31.3) 81.6 (33.5) 84.9 (30.9) FU1 86.8 (30.2) 85.1 (32.6) 5.1 (1.6 to 11.8) 77.4 (37.7) 82.7 (33.4) 3.9 (11.4 to 3.6) FU2 86.0 (30.9) 87.1 (30.3) 1.3 (6.1 to 8.7) 77.5 (37.9) 74.0 (38.7) 5.5 (4.3 to 15.4) Vitality Baseline 58.6 (21.3) 58.5 (23.3) 53.6 (21.1) 54.9 (21.3) FU1 60.8 (21.7) 61.7 (19.4) 0.5 (3.7 to 4.7) 57.1 (22.3) 55.7 (20.3) 2.0 (2.3 to 6.3) FU2 61.6 (20.7) 63.9 (19.0) 1.4 (5.7 to 2.9) 55.3 (22.5) 58.3 (21.1) 1.8 (7.0 to 3.4) Bodily pain Baseline 76.3 (23.9) 76.6 (25.1) 72.3 (25.4) 75.5 (25.0) FU1 78.6 (24.7) 77.4 (24.4) 2.3 (2.4 to 7.1) 75.8 (24.6) 75.8 (26.0) 0.2 (5.3 to 4.9) FU2 78.2 (24.5) 76.1 (23.8) 1.7 (3.8 to 7.2) 74.9 (24.9) 75.2 (19.4) 0.2 (6.8 to 6.4) General health Baseline 73.5 (19.8) 73.4 (18.9) 66.0 (20.6) 71.2 (20.0) FU1 75.2 (20.7) 74.9 (18.4) 0.8 (2.5 to 4.0) 67.9 (21.4) 69.9 (20.7) 1.0 (2.4 to 4.3) FU2 75.0 (18.6) 73.7 (18.5) 1.7 (2.0 to 5.4) 68.6 (21.5) 72.5 (19.4) 0.03 (3.9 to 3.8) a b c Higher score indicates better health state (range 0 – 100). Adjusted for baseline. Larger than expected difference due to baseline imbalance of responders at FU1. Table 4 Respondents indicating ‘Strongly Agree/ Agree’ with statements on causes of breast cancer, n (%) Grampian Wales a a Nurse counsellor Geneticist P-value Nurse counsellor Geneticist P-value Stress is a major cause of breast cancer Baseline 68 (36) 40 (42) 68 (35) 34 (35) Immediately after episode 80 (47) 39 (48) 0.98 63 (38) 31 (37) 0.98 Six months after episode 71 (42) 31 (44) 0.83 55 (37) 26 (36) 0.91 One close relative with breast cancer always increases your risk considerably Baseline 157 (83) 77 (81) 165 (84) 80 (83) Immediately after episode 127 (73) 52 (63) 0.12 130 (77) 65 (77) 0.99 Six months after episode 123 (77) 56 (76) 0.97 122 (81) 58 (80) 0.88 Minor injury can cause breast cancer Baseline 37 (20) 20 (21) 62 (32) 23 (24) Immediately after episode 38 (22) 22 (27) 0.50 50 (30) 16 (19) 0.09 Six months after episode 47 (29) 17 (23) 0.39 50 (33) 18 (25) 0.24 a 2 From w test (Yates’ corrected – for 2 2 table). scale, for Grampian and Wales, respectively. The data suggested Patient knowledge ‘equivalence’ in these outcomes, in both trials, at both follow-up points, with the exception of the mental health score at the second Table 4 summarises the proportions of participants who agreed or follow-up point in the Grampian trial, which indicated that strongly agreed with each statement on a possible cause of breast ‘uncertain equivalence’. cancer. The results are generally similar between the two trial Table 3 summarises the SF36 scores for the other health status locations, and between intervention and control arms within each outcomes. Scores were generally high, the lowest being observed trial. Misunderstanding was greatest for the effect of having a first- for the vitality domain in both trials. On average, higher scores degree relative with breast cancer. No consistent, or statistically were observed in all domains in the Grampian trial compared with significant improvements in knowledge were observed for any of the Wales trial. the three notional risk factors. British Journal of Cancer (2006) 95(4), 435 – 444 & 2006 Cancer Research UK Genetic nurse counsellors can be an acceptable and cost-effective N Torrance et al & 2006 Cancer Research UK British Journal of Cancer (2006) 95(4), 435 – 444 Grampian FU1 (immediately after counselling episode) FU2 (6 months later) Given information wanted about breast cancer causes Given information wanted about breast cancer causes Given information wanted about breast cancer risks Given information wanted about breast cancer risks Given information wanted about breast cancer prevention Given information wanted about breast cancer prevention Given information wanted about genetic tests Given information wanted about genetic tests The doctor/nurse listened Concerns taken seriously Concerns taken seriously The doctor/nurse listened Satisfied with waiting time for first appointment Satisfied with waiting time for first appointment The consultation was helpful The consultation was helpful Satisfied overall Nurse counsellor 0 20406080 100 Satisfied overall Geneticist % Agree/strongly agree Nurse counsellor 0 20406080 100 Geneticist % Agree/strongly agree Wales FU1 (immediately after counselling episode) FU2 (6 months later) Given information wanted about breast cancer causes Given information wanted about breast cancer causes Given information wanted about breast cancer risks Given information wanted about breast cancer risks Given information wanted about breast cancer prevention Given information wanted about breast cancer prevention Given information wanted about genetic tests Given information wanted about genetic tests Concerns taken seriously The doctor/nurse listened The doctor/nurse listened Satisfied with waiting time for first appointment Satisfied with waiting time for first appointment Concerns taken seriously The consultation was helpful The consultation was helpful Satisfied overall Satisfied overall Nurse counsellor Nurse counsellor 0 20406080 100 0 20406080 100 Geneticist Geneticist % Agree/strongly agree % Agree/strongly agree Figure 2 Patient satisfaction with genetic counselling. Clinical Studies Clinical Studies Genetic nurse counsellors can be an acceptable and cost-effective N Torrance et al Patient satisfaction DISCUSSION High levels of patient satisfaction were observed in both trials. This study suggests that, for the initial episode of genetic Figure 2 shows the views of patients on the specific aspects of counselling for risk of breast cancer, nurse counsellors can provide satisfaction with services immediately following the genetic care that appears to be equivalent to that provided by clinical counselling episode and at the 6 month follow-up point. geneticists, in terms of patients’ psychosocial outcomes and satisfaction. The similar findings in the two separate trials support the generalisability of the findings. Acceptability to referring GPs Response rates to the baseline and follow-up surveys were high and we were able to assess the stability of the outcomes over a 6- In all, 74 and 87 GPs in Grampian and Wales, respectively, referred month period following the counselling episode. A high proportion at least one patient who was randomised to the nurse counsellor of eligible patients were recruited into both trials, suggesting that and who attended the genetic clinic. In all, 68 GPs in Grampian our study populations were representative of the target popula- (response rate 92%) and 75 GPs in Wales (response rate 86%) tions. The small but distinct baseline differences between the two responded to the acceptability survey. Grampian respondents, 60 sets of trial participants probably reflect real differences in the (88%) and 52 (69%) Wales respondents could not differentiate patient populations from which they were drawn. Participants in whether their patient had been seen by a nurse counsellor or a Wales had slightly higher anxiety levels overall than the Grampian clinical geneticist. Almost all respondents (Grampian 100%, Wales participants, but these were balanced between intervention and 98.7% (n¼ 74)) reported that they would be happy for future control arms within each setting. After genetic counselling, small referred patients to be seen at the clinic by the genetic nurse reductions in anxiety levels were seen in all groups. Even though counsellor. Overall satisfaction with the medical genetics service the participants’ baseline mean anxiety scores were slightly higher was high with 91% (62 out of 68) GPs in Grampian and 89% (67out than the norm for adult women (score of 35), they were of 75). In Wales reporting that they were ‘very satisfied/satisfied’ comparable to anxiety levels found in other studies that have with the service provided by the respective medical genetics used the STAI (Cull et al, 1998; Cull et al, 1999; Julian-Reynier services. et al, 1999; Brain et al, 2000). The between-trial differences support the decision to conduct two parallel trials rather than multicentre trial, where the data would have been pooled (Bowling, 1997). Economic evaluation The unit costs per counselling appointment Comparative data for SF-36 scores for these two general were similar for the clinical geneticist arms in both trials. Cost geographical populations (Garratt et al, 1993; Lyons et al, 1995) differences between intervention and control arms across the two also suggest that our trial participants were slightly more anxious locations were largely driven by staff costs. For Grampian, the than the underlying populations, but the between-trial differences marginal (additional) cost per single counselling appointment was were similar to the background between-population differences. It d17.98 higher for the control compared with the intervention arm; is also possible that these apparent differences reflect the different in Wales, the marginal cost per counselling appointment was healthcare referral processes in operation in the two locations. For d12.50 higher for the intervention than the control arm. example, in Wales, women who were considered to be at increased Table 5 presents the health service costs per counselling episode. familial risk at the breast screening clinic (Breast Test Wales) were For Grampian women randomised to the nurse counsellor, the first referred to a surgeon who, after review, made the referral for ‘counselling episode’ ranged from 1 to 4 appointments. Most genetic counselling. In Grampian, many women were referred women received one (149 out of 193, 77%) or two (38, 20%) more directly by GPs or other providers to the genetics clinic. It is appointments. In the control arm, 81 out of 96 (84%) participants possible that these differences in referral pathways contributed to received one, and 13 (13.5%) participants received two, appoint- ments. The mean cost per patient for the counselling episode was different anxiety levels in women by the time they received a d136.55 for the nurse counsellor arm and d148.30 for the clinical genetics clinic appointment, and also influenced their risk perceptions. Despite these apparent population differences, the geneticist arm, a difference of d11.54 (95% CI, d25.43, 1.94). In effectiveness data from the two trials were very similar. Wales all participants received one appointment only, hence the Few of the participants in these trials would have been eligible cost per episode and appointment were the same. for genetic testing, and this was not the focus of the study (which Overall, the cost per counselling episode for the nurse counsellor was the period in which initial assessments were made). In both was d3.55 lower in Grampian compared with Wales, and for the locations, women who were assessed as suitable for genetic testing geneticist arm was d20.70 higher in Grampian than in Wales. would have been referred to a consultant geneticist for follow-up Health service unit costs in both centres appeared sensitive to care. Our study end point was the point at which a woman learnt of grades of staff employed (nurses and doctors), level of supervision her risk status, and further follow-up or management arrange- of nurse specialists, and length of counselling appointments, but ments were made. We do not have access to data on how many of not to choice of discount rate or lifespan of equipment items. Table 5 Comparison of health service costs per counselling episode Grampian Wales Mean number of randomised Unit Mean total cost per Mean number of randomised Unit Mean total cost per a a Group appointments (range) cost patient (d) appointments (range) cost patient (d) Nurse counsellor 1.26 (1 – 4) 108.01 136.55 1 (1 – 1) 140.10 140.10 Geneticist 1.18 (1 – 3) 125.99 148.30 1 (1 – 1) 127.60 127.60 Difference in cost 11.54 +12.50 (nurse counsellor – geneticist) a b Unit cost of counselling appointment. (95% CI, d25.43, d1.94). British Journal of Cancer (2006) 95(4), 435 – 444 & 2006 Cancer Research UK Genetic nurse counsellors can be an acceptable and cost-effective N Torrance et al the participants actually went ahead with genetic testing, or their assessment service include the concept of working in liaison or actual test results. outreach settings (Emery et al, 1999; Fry et al, 1999). Two The primary outcome measure was anxiety, the reduction of which randomised controlled trials have reported on different models of is regarded as a key counselling objective (Shaw et al, 1999; Brain service delivery for genetic counselling for risk of breast cancer et al, 2000; Meiser and Halliday, 2002) and a number of evaluations (Brain et al, 2000; Fry et al, 2003). In the trial by Brain et al (2000), of genetic counselling for familial cancer have identified the pre- and the addition of specialist genetic assessment to the standard postcounselling assessment of generalised anxiety as a main outcome surgical consultation had no effect on patients’ psychological measure (Cull et al, 1998, 1999; Julian-Reynier et al, 1999; Brain et al, outcomes, risk perception or satisfaction, although knowledge of 2000; Kent et al, 2000; Bish et al, 2002; Bowen et al,2004).Evidence cancer genetics showed greater improvement. The other RCT from systematic reviews suggests that, overall, genetic counselling found community-based genetic nurse specialists to be generally has the effect of significantly reducing patients’ anxiety levels, at least comparable to the standard service (consultant geneticist), in in the short-term (Meiser and Halliday, 2002; Butow et al, 2003; terms of psychosocial outcomes and patient satisfaction, with the Braithwaite et al, 2004). The apriori equivalence limits for the additional benefit of lower staff and patient costs (Fry et al, 2003; primary outcome (STAI) were set at a very strict level, and in reality Holloway et al, 2004). they probably represent a smaller difference than would normally be There is now an emerging body of research evidence on the considered clinically significant between two clinicians considered relative effectiveness of nurses in specialist roles compared with equally competent; however, results showed that some outcomes doctors in both primary care (Horrocks et al, 2002), and secondary were considered ‘equivalent’ even at this strict level. care settings (Kinley et al, 2001; Sharples et al, 2002), with most Three common mistaken beliefs on the causes of breast cancer studies finding that nurses working to guidelines appear to provide were not influenced by counselling; the most persistent and care that is equal to that provided by doctors, with comparable erroneously held belief was in the influence of one close relative health outcomes. Previous randomised trials in secondary care with breast cancer on a person’s own risk. This may have reflected have found nurse practitioners to be either cost neutral (Kinley a general misconception, or reflect the participants’ own personal et al, 2001) or more expensive than doctors either because of salary risk perceptions. The knowledge we assessed was not specific to costs (Sakr et al, 2003) or greater resource use (Sharples et al, cancer genetics, and our findings are consistent with genetic 2002). In this study, we found that the relative costs of the nurse counselling delivered by either doctors or nurses being equally counsellors, compared with the doctor-led service, depended on effective (or ineffective) in educating patients about breast cancer the grade of medical staff whose time was being replaced by the risk. Braithwaite et al (2004) suggest that genetic counselling can nurse counsellor, and the extent of supervision required – of both be effective in improving knowledge related to breast cancer nurse counsellors and less experienced medical staff. Surprisingly, genetics, compared with no counselling or counselling delivered by the sensitivity analysis did not suggest that nurse counsellors of nongenetics specialists. However, the observed level of misunder- lower grades would be less costly than the nurses employed here, standing seen in this study suggests that more effective population on the basis that the cost saving in lower salaries would be offset by interventions are required to improve general knowledge of breast the greater need for consultant supervision for nurses at lower cancer risks. grades. Patient satisfaction with information provided, staff attitudes The results of this study add to the emerging body of evidence and the overall clinic procedures were high overall in both trials supporting the effectiveness, and possible cost-effectiveness, of irrespective of randomised group. Similar levels of patient nurse counsellors working under the supervision of consultant satisfaction have been reported in other trials of genetic geneticists, and should be taken into account by decision-makers counselling services for risk of breast cancer (Brain et al, 2000; planning and evaluating genetics health services. Hopwood et al, 2004; Holloway et al, 2004). In addition, the acceptability of the genetic nurses counsellors was high among referring GPs in both trial settings. However, it is possible that the ACKNOWLEDGEMENTS difference in service model (i.e. appointment of the nurse counsellors) may be imperceptible at primary care level as We thank Shelley Farrar, Richard Gordon, Tracy Ibbotson and individual GPs referred only one or two patients during the trial. Leslie Walker who were involved in the original grant application The lack of well-designed and rigorously conducted randomised and study design. Thanks go to the genetic nurse counsellors who controlled trials, with reporting to CONSORT standards, in the were involved in the study, and to all the NHS clinical and field of service delivery for genetic counselling for familial cancer administrative staff in both clinical genetics centres. We are has been observed in a number of recent publications (Butow et al, grateful to the patients who participated in both trials. This study 2003; Braithwaite et al, 2004; Hopwood et al, 2004; NICE, 2004). was funded by the NHS Health Technology Assessment pro- Alternative models for how nurse counsellors might provide a risk gramme. REFERENCES Brain K, Gray J, Norman P, France E, Anglim C, Barton G, Parsons E, Bowen DJ, Burke W, McTiernan A, Yasui Y, Andersen MR (2004) Breast Clarke A, Sweetland H, Tischkowitz M, Myring J, Stansfield K, Webster cancer risk counseling improves women’s functioning. Patient Edu D, Gower-Thomas K, Daoud R, Gateley C, Monypenny I, Singhal H, Couns 53: 79– 86 Branston L, Sampson J, Roberts E, Newcombe R, Cohen D, Rogers C, Bowling A (1997) Research Methods in Health: investigating health and Mansel R, Harper P (2000) Randomized trial of a specialist genetic health services. Buckingham: Open University Press assessment service for familial breast cancer. J Natl Cancer Inst 92(16): Butow PN, Lobb EA, Meiser B, Barratt A, Tucker KM (2003) Psychological 1345– 1351 outcomes and risk perception after genetic testing and counselling in Braithwaite D, Emery J, Walter F, Prevost AT, Sutton S (2004) Psychological breast cancer: a systematic review. Med J Aust 78: 77– 81 impact of genetic counselling for familial cancer: a systematic review and Claus EB, Risch N, Thompson WD (1994) Autosomal dominant inheritance meta-analysis. J Natl Cancer Inst 96(20): 122– 133 of early onset breast caner. Cancer 73: 643–651 Bish A, Sutton S, Jacobs C, Levene S, Ramirez A, Hodgson S (2002) Changes Cull A, Anderson EDC, Campbell S, Mackay J, Smyth E, Steel CM (1999) in psychological distress after cancer genetic counselling: a comparison The impact of genetic counselling about breast cancer risk on women’s of affected and unaffected women. Br J Cancer 86: 43–50 risk perceptions and levels of distress. Br J Cancer 79(3/4): 501 –508 & 2006 Cancer Research UK British Journal of Cancer (2006) 95(4), 435 – 444 Clinical Studies Clinical Studies Genetic nurse counsellors can be an acceptable and cost-effective N Torrance et al Cull A, Miller H, Porterfield T, Mackay J, Anderson ED, Steel CM, Elton RA gical morbidity and health beliefs in women attending for genetic (1998) The use of videotaped information in cancer genetic counselling: counselling. Br J Cancer 74: 482–487 A randomised evaluation study. Br J Cancer 77(5): 830–837 Lyons RA, Fielder H, Littlepage BN (1995) Measuring health status with the Emery J, Watson E, Rose P, Andermann A (1999) A systematic review of the SF-36: the need for regional norms. J Public Health Med 17: 46–50 literature exploring the role of primary care in genetic services. Fam Marteau TM, Bekker HB (1992) The development of a six-item short-form Pract 16: 426–445 of the state scale of the Spielberger State-Trait Anxiety Inventory (STAI). Fry A, Campbell H, Gudmunsdottir H, Rush R, Porteous M, Gorman D, Cull Br J Clin Psychol 31: 301–306 A (1999) GPs’ views on their role in cancer genetics services and current Marteau TM, Walker P, Giles J, Smail M (1990) Anxieties in women practice. Fam Pract 16: 468–474 undergoing colposcopy. Br J Obstet Gynaecol 97: 859 –861 Fry A, Cull A, Appleton S, Rush R, Holloway S, Gorman D, Cetnarskyj R, Meiser B, Halliday JL (2002) What is the impact of genetic counselling in Thomas R, Campbell J, Anderson E, Steel M, Porteous M, Campbell H women at increased risk of developing hereditary breast cancer? A meta- (2003) A randomised trial of breast cancer genetics services in South East analytic review. Soc Sci Med 54: 1463–1470 Scotland: psychological impact. Br J Cancer 89: 653– 659 Miedzybrodzka ZH, Hall MH, Mollison J, Templeton A, Russell IT, Dean JC, Garratt AM, Ruta DA, Abdalla MI, Buckingham JK, Russell IT (1993) The Kelly KF, Marteau TM, Haites NE (1995) Antenatal screening for carriers SF36 health survey questionnaire: an outcome measure suitable for of cystic fibrosis: randomised trial of stepwise v couple screening. BMJ routine use within the NHS? BMJ 306: 1440 –1444 310: 353–357 Holloway S, Porteous M, Cetnarskyj R, Anderson E, Rush R, Fry A, Gorman National Institute for Clinical Excellence (2004) Clinical Guidelines and D, Steel M, Campbell H (2004) Patient satisfaction with two different Evidence Review for the Classification and Care of Women at Risk of models of cancer genetic services in south-east Scotland. Br J Cancer 90: Familial Breast Cancer. London: National Collaborating Centre for 582–589 Primary Care/University of Sheffield, National Institute for Clinical Hopwood P, Wonderling D, Watson M, Cull A, Douglas F, Cole T, Eccles D, Excellence (www.nice.org.uk) Gray J, Murday V, Steel M, Burn J, McPherson K (2004) A randomised Office for National Statistics (2005) http://www.statistics.gov.uk/cci/ comparison of UK genetic risk counselling services for familial cancer: nugget.asp?id¼ 575 [accessed 08/12/2005] psychosocial outcomes. Br J Cancer 91(5): 884 –892 Pocock SJ (1983) Clinical trials: a practical approach. Chichester: Wiley Horrocks S, Anderson E, Salisbury C (2002) Systematic review of whether Sakr M, Kendall R, Angus J, Saunders A, Nicholl J, Wardrope J (2003) nurse practitioners working in primary care can provide equivalent care Emergency nurse practitioners: a three part study in clinical and cost to doctors. BMJ 324: 819 –823 effectiveness. Emerg Med J 20: 158– 163 Jones B, Jarvis P, Lewis JA, Ebbutt AF (1996) Trials to assess equivalence: Sharples LD, Edmunds J, Bilton D, Hollingworth W, Caine N, Keogan M, the importance of rigorous methods. BMJ 313: 36–39 Exley A (2002) A randomised controlled crossover trial of nurse Julian-Reynier C, Eisinger F, Chabal F, Aurran Y, Bignon YI, Machelard- practitioner versus doctor led outpatient care in a bronchiectasis clinic. Roumagnac M, Maugard C, Nogues C, Vennin P, Sobol H (1999) Cancer Thorax 57: 661–666 genetic consultation and anxiety in healthy consultees. Psychol Health 14: Shaw C, Abrams K, Marteau TM (1999) Psychological impact of predicting 379–390 individuals’ risks of illness: a systematic review. Soc Sci Med 49: 1571– Kent G, Howie H, Fletcher M, Newbury-Ecob R, Hosie K (2000) The 1598 relationship between perceived risk, thought intrusiveness and emotional Shiloh S, Avdor O, Goodman RM (1990) Satisfaction with genetic well-being in women receiving counselling for breast cancer risk in a counseling: dimensions and measurement. Am J Med Genet 37: 522–529 family history clinic. Br J Health Psychol 5: 15– 26 Skirton H, Barnes C, Guilbert P, Kershaw A, Kerzin-Storrar L, Patch C, Kinley H, Czoski-Murray C, George S, McCabe C, Primrose J, Reilly C, Curtis G, Walford-Moor J (1998) Recommendations for education and Wood R, Nicolson P, Healy C, Read S, Norman J, Janke E, Alhameed H, training of genetic nurses and counsellors in the United Kingdom. J Med Fernandez N, Thomas E (2001) Extended scope of nursing practice: a Genet 35: 410–412 multicentre randomised controlled trial of appropriately trained nurses Ware JE, Kosinski M, Keller SD (1994) SF36 Physical and mental and pre-registration house officers in pre-operative assessment in component summary measures: a users manual. The Health Institute: elective general surgery. Health Technol Assess 5(20): 1–87 Boston Kinmonth AL, Reinhard J, Bobrow M, Pauker S (1998) The new genetics. Watson M, Lloyd S, Davidson J, Meyer L, Eeles R, Ebbs S, Murday V (1999) Implications for clinical services in Britain and the United States. BMJ The impact of genetic counselling on risk perception and mental health 316: 67–70 in women with a family history of breast cancer. Br J Cancer 79(5 –6): Lerman C, Daly M, Sands C, Balshem A, Lustbader E, Heggan T, Goldstein 868–874 L, James J, Engstrom P (1993) Mammography adherence and Wonderling D, Hopwood P, Cull A, Douglas F, Watson M, Burn J, psychological distress among women at risk for breast cancer. J Natl McPherson K (2001) A descriptive study of UK cancer genetics services: Cancer Inst 85: 1074– 1080 an emerging clinical response to the new genetics. Br J Cancer 85: 166–170 Lloyd S, Watson M, Waites B, Meyer L, Eeles R, Ebbs S, Tyle A (1996) Zigmond AS, Snaith RP (1983) The hospital anxiety and depression scale. Familial breast cancer: a controlled study of risk perception, psycholo- Acta Psychiatr Scand 67: 361–370 British Journal of Cancer (2006) 95(4), 435 – 444 & 2006 Cancer Research UK http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png British Journal of Cancer Springer Journals

Genetic nurse counsellors can be an acceptable and cost-effective alternative to clinical geneticists for breast cancer risk genetic counselling. Evidence from two parallel randomised controlled equivalence trials

Loading next page...
 
/lp/springer-journals/genetic-nurse-counsellors-can-be-an-acceptable-and-cost-effective-AWia0Jabac

References (48)

Publisher
Springer Journals
Copyright
Copyright © 2006 by The Author(s)
Subject
Biomedicine; Biomedicine, general; Cancer Research; Epidemiology; Molecular Medicine; Oncology; Drug Resistance
ISSN
0007-0920
eISSN
1532-1827
DOI
10.1038/sj.bjc.6603248
Publisher site
See Article on Publisher Site

Abstract

British Journal of Cancer (2006) 95, 435 – 444 & 2006 Cancer Research UK All rights reserved 0007 – 0920/06 $30.00 www.bjcancer.com Genetic nurse counsellors can be an acceptable and cost-effective alternative to clinical geneticists for breast cancer risk genetic counselling. Evidence from two parallel randomised controlled equivalence trials ,1 1 2 3 4 4 5 5 N Torrance , J Mollison , S Wordsworth , J Gray , Z Miedzybrodzka , N Haites , A Grant , M Campbell , 1 3 6 MS Watson , A Clarke and B Wilson 1 2 Department of Public Health, Medical School, University of Aberdeen, Foresterhill, Aberdeen AB25 2ZD, UK; Health Economics Research Centre, 3 4 University of Oxford, Oxford OX3 7LF, UK; Institute of Medical Genetics, University Hospital of Wales, Cardiff CF14 4XN, UK; Department of Medicine & Therapeutics, University of Aberdeen, Aberdeen AB25 2ZD, UK; Health Services Research Unit, University of Aberdeen, Aberdeen AB25 2ZD, UK; Department of Epidemiology & Community Medicine, University of Ottawa, 451 Smyth Road, Ottawa, Ontario, Canada K1H 8M5 This study compared genetic nurse counsellors with standard services for breast cancer genetic risk counselling services in two regional genetics centres, in Grampian region, North East Scotland and in Cardiff, Wales. Women referred for genetic counselling were randomised to an initial genetic counselling appointment with either a genetic nurse counsellor (intervention) or a clinical geneticist (current service, control). Participants completed postal questionnaires before, immediately after the counselling episode and 6 months later to assess anxiety, general health status, perceived risk and satisfaction. A parallel economic evaluation explored factors influencing cost-effectiveness. The two concurrent randomised controlled equivalence trials were conducted and analysed separately. In the Grampian trial, 289 patients (193 intervention, 96 control) and in the Wales trial 297 patients (197 intervention and 100 control) returned a baseline questionnaire and attended their appointment. Analysis suggested at least likely equivalence in anxiety (the primary outcome) between the two arms of the trials. The cost per counselling episode was d11.54 less for nurse-based care in the Grampian trial and d12.50 more for nurse-based care in Cardiff. The costs were sensitive to the grade of doctor (notionally) replaced and the extent of consultant supervision required by the nurse. In conclusion, care based on genetic nurse counsellors was not significantly different from conventional cancer genetic services in both trial locations. British Journal of Cancer (2006) 95, 435–444. doi:10.1038/sj.bjc.6603248 www.bjcancer.com Published online 11 July 2006 & 2006 Cancer Research UK Keywords: genetic counselling; familial cancer; breast cancer; randomised control trial; equivalence; anxiety Familial cancer is an area in which genetic knowledge has to increased lifetime risks of developing the disease, often at an progressed rapidly over the past 10 years, and where patient earlier age than sporadic disease (Claus et al, 1994). However, demand for genetics services for information, counselling and these mutations are implicated in around only 5–10% of breast mutation testing has increased dramatically. Even with the cancer cases, with the result that genetic testing is not appropriate introduction of referral guidelines, regional genetics clinics are for the majority of women with a family history of breast cancer. facing the challenge of increasing demand within fixed resources Genetic counselling services aim to identify individuals who have a (Kinmonth et al, 1998), and these difficulties may be exacerbated significantly increased genetic risk of cancer and counsel them as new mutations are identified and further demands are made on about appropriate risk management to reduce morbidity and genetics clinics. mortality (Fry et al, 2003). Access to surveillance, testing and other Breast cancer is the most common cancer affecting UK women, interventions is usually dependent on the outcome of the initial resulting in a lifetime risk of one in nine of developing the disease genetic counselling and risk assessment. (Office for National Statistics, 2005). The discovery of breast Recent research in the area of breast cancer genetic risk cancer susceptibility mutations has attracted widespread publicity counselling has generally concentrated on women’s psychological and women with a family history of breast cancer dominate status and risk perception. Evidence from systematic reviews referrals throughout the UK for any cancer genetic risk counselling suggests that genetic counselling does not appear to have any (Wonderling et al, 2001; Hopwood et al, 2004). Mutations in the adverse effect on psychological outcomes (Meiser and Halliday, known inherited susceptibility genes, BRCA1 and BRCA2, give rise 2002; Butow et al, 2003; Braithwaite et al, 2004), and may convey some short-term benefits in decreasing general anxiety and in improving the accuracy of women’s perceptions of their personal *Correspondence: Dr N Torrance; E-mail: n.torrance@abdn.ac.uk risk. Historically, clinical genetic risk assessments in the UK have Received 31 March 2006; revised 5 June 2006; accepted 8 June 2006; published online 11 July 2006 been conducted largely by medical doctors, with a variable amount Clinical Studies Clinical Studies Genetic nurse counsellors can be an acceptable and cost-effective N Torrance et al of support for genetic counselling provided by genetic nurses and Eligible patients were randomised 2 : 1 to intervention: control counsellors, social workers and psychologists as well. Many groups. This uneven allocation was chosen to give greater specialist centres have recently begun to expand their genetic risk experience of the novel intervention and was taken account of in counselling capacity by recruiting genetic nurses or associates the sample size calculation. Postal questionnaires were adminis- (Skirton et al, 1998). This trend is likely to continue. However, tered to participants at baseline (before the clinic appointment), there is a lack of formal evaluation of the effectiveness of this immediately following the counselling episode (FU1), and 6 strategy and the lack of well designed and conducted randomised months later (FU2). controlled trials of genetic service models has been noted in a number of recent publications (Butow et al, 2003; Braithwaite et al, Standard care in two trial locations (clinical geneticist) 2004; Hopwood et al, 2004; NICE, 2004). We therefore conducted a pragmatic evaluation comparing As indicated above, the control arms reflected current standard of genetic nurse counsellors as an alternative to a physician-based care in each location, and no attempt was made to alter this. The service (standard care) in assessing patients newly referred for process for patients referred for genetic assessment was similar in genetic counselling for risk of breast cancer. We focused on both centres: family history taking, pedigree construction, patient-centred outcomes, and acceptability to referring general confirmation of cancer diagnoses in affected relatives, risk practitioners, and conducted a concurrent economic evaluation. In assessment, genetic counselling with advice on preventive order to increase the generalisability of our findings, we conducted strategies and, if appropriate screening, prophylactic surgery, the evaluation as two randomised controlled equivalence trials and/or genetic testing. Initially, the clinical geneticist took the with differing patient population and service characteristics. family history, constructed the pedigree, confirmed cancer diagnoses, etc in both centres. In Grampian, clerical and nursing support was available to enable the family tree to be constructed in advance of the clinic appointment. In Wales the clinical geneticist MATERIALS AND METHODS did all of the preclinic preparation. In both settings, clinic appointments involved a clinical geneticist and lasted approxi- Participants and procedures mately 45 min. Two concurrent randomised controlled trials were conducted to assess the equivalence of care provided by a genetic nurse counsellor with clinical geneticist-based (standard) care. The Intervention (genetic nurse counsellors) study locations were the Grampian University Hospitals NHS In both trials, the nurse counsellor made the initial contact with Trust Clinical Genetics Service (in Aberdeen) and the Wales the patient. Each followed the format of the standard care Genetics Service (in Cardiff). The study was designed with a described above for their own centre, with a nurse counsellor pragmatic approach; that is, the goal was to gain an insight into the substituting for clinical geneticist input. Each clinic was therefore overall effects of substituting a nurse counsellor-based approach free to make its own organisational arrangements for the nurse-led for a physician-based approach, while accepting that contextual counselling arm, so long as there were adequate minimum factors inevitably vary between health care settings. Thus, there supervision arrangements, a clinical geneticist was professionally were differences in regular (doctor-based) care in the two trial and managerially accountable for the patient care delivered by a locations, but both represented an acceptable ‘standard of care’ for nurse counsellor, and the same risk assessment algorithms were breast cancer genetic risk assessment and counselling within the followed within each clinic (by both nurse counsellors and clinical NHS. Similarly, the intervention arms reflected nurse counsellor- geneticists). The focus of the study was the ‘counselling episode’: based care developed and managed according to local protocols the time from appointment scheduling until the patient was either and supervisory arrangements. As a clinical geneticist was formally discharged or further management arrangements were made. This responsible in each location for patient care, whether it was provided the ‘boundary’ to the intervention – all activities from delivered by a doctor or a nurse counsellor, we were not concerned this point on were part of the ‘usual’ cancer genetics service. with assessing performance of the nurse counsellors in relation to Participants who were considered likely candidates for genetic technical procedures such as pedigree construction or calculation testing (i.e. high risk) were referred for further follow-up with a of risk estimates. Rather, we concentrated on other outcomes of consultant geneticist. In both trials, a consultant geneticist actively interest, including provider and patient satisfaction. By imple- supervised the nurse counsellors, meeting with them on a weekly menting trial procedures rigorously, collecting identical data sets, basis. Training was delivered according to the needs of the nurse but analysing the two trials separately, we were able to promote counsellors when they were appointed. internal validity within each trial but also comment on external validity (i.e. generalisability) (Pocock, 1983). Patients were eligible if they were newly referred because of Data collection concern about family history of breast cancer, were aged 18 years or over, and literate in English. Patients previously affected by Sociodemographic data Sociodemographic data were collected breast cancer were included. Patients were excluded if they had from the women in the baseline questionnaire, and included age, previously attended the genetics clinic or were known to be a marital status, number of children, ethnicity and educational level. member of a family in which a BRCA1 or BRCA2 mutation had been previously identified. Psychological outcomes The primary outcome was patient Eligible patients were identified from referral letters and consent anxiety, measured using the six-item short-form of the state scale to contact was obtained from each patient’s GP. Before their clinic of the Spielberger State-Trait Anxiety Inventory (STAI; Marteau appointment, patients were sent a letter inviting their participa- and Bekker, 1992). The short version produces result comparable tion, an information sheet explaining the study and a consent with the full state scale and has been used in other genetic form. Those who consented were randomised to the intervention counselling research (Miedzybrodzka et al, 1995; Bish et al, 2002). (genetic nurse counsellor-based care) or control group (clinical We also included the Hospital Anxiety and Depression Scale geneticist-based care). The random allocation schedule sequence (HADS; Zigmond and Snaith, 1983), which gives separate measures for each trial was computer-generated and concealed within a of anxiety and depression assessed over the past week (range of Microsoft Access database. To avoid possible contamination, scores from 0 to 21, seven items for anxiety and seven for members of the same family were allocated to the same trial group. depression). British Journal of Cancer (2006) 95(4), 435 – 444 & 2006 Cancer Research UK Genetic nurse counsellors can be an acceptable and cost-effective N Torrance et al Health-related quality of life The Short Form 36 (SF-36) health but before data were released for analysis. For anxiety as measured survey instrument was used to measure perception of general by HADS, and the role-emotional and mental health domains of health status (Ware et al, 1994). It measures patients’ rating of the SF36, we defined the equivalence limit as one-third of a their own health status in eight areas or domains: physical standard deviation of the baseline scores. functioning, role physical, role emotional, social functioning, We analysed data using the ‘intention to treat’ principle, in mental health, vitality, bodily pain and general health perception. order to minimise the effect of selection bias and tested the For each dimension item scores range from 0 (worst possible robustness of the findings by comparing the findings to a health state) to 100 (best possible health state). ‘treatment-received’ analysis. Analysis was performed using SPSS v11.0. Perceived risk of breast cancer We assessed participants’ perceived risk of breast cancer at all three time points with an Sample size The primary outcome was anxiety as assessed by item used in previous research (Lerman et al, 1993; Lloyd et al, STAI (short form), with an equivalence limit of 74 units. We 1996; Watson et al, 1999; Brain et al, 2000). Women were asked to required 214 participants in the intervention group and 107 in the assess their own risk of developing breast cancer relative to a control group to allow detection of equivalence (two sided), with notional ‘average’ woman, on a five-point scale (‘much lower/ 80% power at the 5% significance level. This assumed a standard lower/about the same/higher/much higher’). Women previously deviation of 12 units for STAI (Marteau et al, 1990), and affected by breast cancer were asked to skip this section of the incorporated the 2 : 1 allocation ratio. questionnaire to avoid potential confusion. Economic evaluation Knowledge of breast cancer risk factors We assessed participants’ understanding of risk factors for breast cancer at all three time A cost analysis was conducted alongside the trials, adopting a points. Women were asked how strongly they agreed with three societal perspective to include health service and patient costs. The statements on specific causes of breast cancer (stress, having a ‘cost per patient counselling episode’ for each woman in both trials first-degree relative with breast cancer and minor injury). We used was calculated. Data on health service resource use (staff, a five-point scale (strongly disagree/disagree/not sure/agree/ consumables, rooms and equipment) were collected and local unit strongly agree). For each of these three potential risk factors, costs applied where available, otherwise national rates were used. agreement would indicate a response inconsistent with the Staff time included outpatient appointments, preparation, meet- education content offered during the genetic counselling appoint- ings and time in the nurse counsellor arm for consultant ment. geneticists discussing patient cases with the nurses. The mid- point of salary scales was used, adding employers’ on-costs at 13%. Patient satisfaction and acceptability to referring GPs Patient An equivalent annual cost was calculated for equipment items such satisfaction was measured using a modified version of the as computers, over relevant life-spans for the items, using a 6% Satisfaction with Genetic Counselling Questionnaire developed discount rate. A total cost per counselling episode for each woman by Shiloh et al (1990) and was incorporated into both follow-up was calculated by multiplying the unit cost per single appointment questionnaires. The scale assesses three dimensions of patient by the number of appointments in each trial arm. Sensitivity satisfaction: (1) instrumental (satisfaction with the doctor/nurse’s analysis tested the robustness of the findings and data on patient competence), (2) affective (satisfaction with the doctor/nurse’s time and travel costs were also collected. All costs are reported in personal qualities) and (3) procedural (satisfaction with adminis- Sterling (d) for the price year 2006. trative procedures, such as waiting time and staff conduct). The Approval for the study was obtained from the Joint Ethics acceptability of the genetic nurse counsellors to referring GPs was Committee of Aberdeen University and Grampian Health Board, assessed towards the end of the study. A short questionnaire was and the research ethics committees of Bro Taf and Iechyd used to ascertain whether the referring GPs had noticed any Morgannwg health authorities. difference (including deterioration) in the service provided by the genetics department for their patients, if they would be happy for their patients to be seen by a genetic nurse counsellor in the future RESULTS and an overall rating of their satisfaction with the genetics services. Participants Statistical analysis Grampian In total, 517 referred patients were considered for recruitment (Figure 1). Of the 342 (66%) patients who consented Descriptive statistics were used to describe the study participants. and were randomised (227 intervention, 115 control), 289 (84%) As this was an equivalence trial, a confidence interval (CI) returned a baseline questionnaire and attended clinic (193 (85%) approach was used (Jones et al, 1996). The difference in outcomes intervention, 96 (83%) control). In total, 17 did not receive the was calculated, adjusting for differences in baseline scores using allocated management because of administrative errors (6), joint multiple linear regression. For each outcome, the 95% CIs around family appointments (4) or decision by the head of service (7). the difference were calculated. An outcome was considered ‘equivalent’ when the 95% CI for the difference between the intervention and control group fell completely within a prede- Wales In total, 464 patients were considered for recruitment termined equivalence limit (Jones et al, 1996). Where the 95% CI (Figure 1). Of the 373 (80%) patients who consented and were for the observed difference fell completely outside the equivalence randomised (247 intervention, 126 control), 297 (80%) returned a limit, we considered the outcomes would be ‘nonequivalent’. baseline questionnaire and attended clinic (197 (80%) interven- Where the 95% CI for the observed difference overlapped the tion, 100 (79%) control). Six women did not receive the allocated equivalence limit, the result would be uncertain. For the primary management because of administrative errors (2) and joint family outcome, the STAI score, we defined an a priori strict limit of appointment (4). ‘equivalence’ of 74 units (used in the sample size calculation Baseline data are presented for women who returned a baseline below), and a ‘likely equivalence’ limit of 710 units (which the questionnaire and attended a clinic appointment (Table 1). The trial steering group considered more reflective of actual clinical randomised groups were generally similar in terms of demo- practice). The strict limit was defined before the study started, and graphic characteristics. Within each trial, similar proportions of the likely equivalence limit was determined after the study began, participants allocated to each arm perceived themselves at elevated & 2006 Cancer Research UK British Journal of Cancer (2006) 95(4), 435 – 444 Clinical Studies Genetic nurse counsellors can be an acceptable and cost-effective N Torrance et al British Journal of Cancer (2006) 95(4), 435 – 444 & 2006 Cancer Research UK Progress of participants through the trials Grampian Wales Patients assessed for eligibility (n =464) Patients assessed for eligibility (n =517) Not randomised (n =91) Not randomised (n =175) Reasons: refused to take part (n =26); did not fulfil inclusion Reasons: refused to take part (n =48); did not fulfil inclusion criteria (n =6), no response to recruitment (n =59) criteria (n =28); no response to recruitment (n =99) Randomised (n =373) Randomised (n =342) Allocated to doctor (control) arm (n = Allocated to genetic nurse (intervention) Allocated to genetic nurse (intervention) Allocated to doctor (control) arm (n =115) 126) arm (n =247) arm (n =227) Returned baseline questionnaire and Returned baseline questionnaire and Returned baseline questionnaire and Returned baseline questionnaire and attended clinic (n =96) attended clinic (n =197) attended clinic (n =100) attended clinic (n=193) Received allocated intervention (n =92) Received allocated intervention (n=181) Received allocated intervention (n =191) Received allocated intervention (n =100) Immediate follow-up (n =175) Immediate follow-up (n =83) Immediate follow-up (n =169) Immediate follow-up (n =85) Lost to follow-up (n =13) Lost to follow-up (n =18) Lost to follow-up (n =27) Lost to follow-up (n =14) Withdrew (n =1) Withdrew (n =1) Six-month follow-up (n =163) Six-month follow-up (n =74) Six-month follow-up (n =150) Six-month follow-up (n =73) Lost to follow-up (n =12) Lost to follow-up (n =9) Lost to follow-up (n =19) Lost to follow-up (n =12) Completed trial (n =73) Completed trial (n =150) Completed trial (n =74) Completed trial (n =163) Notes : Notes : a e Did not return baseline questionnaire (n =20), did not attend clinic (n =11), moved away (n =1), Did not return baseline questionnaire (n =39), did not attend clinic (n =10), administrative error (n =1), subsequently found to be not eligible (n =2). subsequently found to be not eligible (n =2). b f Attended appointment with Doctor arm (n =12). Attended appointment with Doctor arm (n =6). c g Did not return baseline questionnaire (n =15), did not attend clinic (n =4). Did not return baseline questionnaire (n =22), did not attendclinic (n =2), subsequently found to be not eligible (n =2). Attended appointment with nurse arm (n =4). Figure 1 Progress of participants through the trials. Clinical Studies Genetic nurse counsellors can be an acceptable and cost-effective N Torrance et al Table 1 Characteristics of study participants at baseline risk, although Wales participants were generally more likely to view themselves as at elevated risk than Grampian participants. Grampian Wales The actual estimated lifetime risk of breast cancer for participants only became available after each initial individual counselling Nurse Nurse episode was complete. In the Grampian trial, participants in the counsellor Geneticist counsellor Geneticist intervention arm were more likely than those in the control arm to be assessed as being at elevated risk, whereas the converse was No. of patients 193 96 197 100 observed for the Wales trial. The analysis by treatment received Age (years): mean 40.7 (10.3) 41.4 (9.4) 39.8 (10.2) 39.0 (9.3) produced only minor differences in results compared with the (s.d.) Married/cohabiting: n 151 (80.3) 74 (77.9) 161 (81.7) 84 (84.8) analysis by intention-to-treat, therefore the data are not reported (%) in this paper. With children: n (%) 157 (83.5) 73 (76.0) 157 (79.7) 73 (73.0) Post-secondary 76 (39.4) 43 (44.8) 73 (37.1) 42 (42.0) education: n (%) Anxiety and general health status Referral source: n (%) Table 2 summarises the primary outcomes for the two trials. There General 133 (68.9) 63 (65.6) 117 (59.4) 56 (56.0) were small but consistent baseline differences between the practitioner Grampian and Wales study populations, but generally scores were Breast surgeon 35 (18.1) 19 (19.8) 73 (37.1) 42 (42.0) comparable between intervention and control arms within each Breast screening 17 (8.8) 10 (10.4) — — trial. With respect to STAI, all adjusted point estimates for the clinic differences between intervention and control groups met the Other 8 (4.1) 4 (4.2) 7 (3.5) 2 (2.0) definition of, at least, ‘likely equivalence’. a a b b For anxiety and depression as measured by HADS, a priori n (%) perceiving 130/181 65/91 (71) 119/ 179 76/90 (84) equivalence limits of one-third of the baseline standard deviation themselves to be at (72) (83) elevated risk at were calculated from the data as 71.4 (Grampian) and 71.5 baseline (Wales) for anxiety and 71.2 (both trials) for depression. These c c d d n (%) assessed at 173/192 74/95 (78) 145/197 84/100 (84) are close to the smallest possible difference in score for an clinic at elevated risk (90) (74) individual, which is 71 point. The results of all analyses were consistent with ‘equivalence’. Also, shown are the observed For perceived risk, participants previously affected by breast cancer are excluded: a b differences in the SF36 role-emotional and mental health domains. Seven nurse counsellor, one geneticist. 10 nurse counsellor, three geneticist. For The equivalence limits were calculated as 711.4 and 713.1 for the assessed risk, participants previously affected by breast cancer are included: Five nurse counsellor, one geneticist. 10 nurse counsellor, three geneticist. role-emotional score, and 76.0 and 76.3 for the mental health Table 2 Psychological outcomes, mean (s.d.) Grampian Wales b b Nurse Equivalence Difference Nurse Equivalence Difference Outcomes counsellor Geneticist limit (95% CI) counsellor Geneticist limit (95% CI) STAI Baseline 37.3 (13.6) 36.5 (12.8) 40.9 (15.1) 40.0 (14.5) Immediately after episode 36.4 (14.0) 34.4 (14.0) 74.0 0.8 (2.1 to 3.7) 38.1 (14.9) 38.9 (15.6) 74.0 1.5 (4.5 to 1.5) Six months after episode 36.0 (13.5) 32.1 (11.7) 2.9 (0.2 to 5.9) 38.9 (14.9) 38.1 (14.1) 0.6 (2.9 to 4.1) HADS Anxiety Baseline 6.7 (4.3) 6.4 (4.5) 8.1 (4.7) 7.4 (4.2) Immediately after episode 6.3 (4.3) 5.5 (3.9) 71.4 0.5 (0.4 to 1.3) 7.0 (4.9) 7.1 (4.8) 71.5 0.4 (1.3 to 0.5) Six months after episode 6.2 (4.4) 5.5 (3.7) 0.1 (0.7 to1.0) 7.4 (4.7) 6.4 (4.1) 0.5 (0.6 to 1.5) HADS Depression Baseline 3.9 (3.7) 3.4 (3.4) 4.5 (3.7) 4.2 (3.8) Immediately after episode 3.5 (3.6) 2.9 (2.8) 71.2 0.3 (0.4 to 1.0) 4.0 (3.8) 4.0 (3.8) 71.2 0.2 (1.0 to 0.5) Six months after episode 3.4 (3.6) 2.8 (2.9) 0.3 (0.5 to 1.0) 4.5 (4.1) 3.9 (3.8) 0.6 (0.4 to 1.5) SF36 Role emotional Baseline 80.5 (34.6) 82.6 (33.4) 74.4 (38.7) 71.0 (40.6) Immediately after episode 81.6 (35.2) 82.5 (33.2) 711.4 1.9 (6.3 to 10.1) 74.8 (39.5) 71.5 (40.0) 713.1 2.9 (6.9 to 12.7) Six months after episode 80.3 (35.9) 86.0 (30.7) 2.5 (11 to 5.9) 74.9 (38.7) 73.1 (42.2) 0.5 (9.4 to 10.5) SF36 Mental health Baseline 71.0 (18.2) 73.6 (17.7) 67.3 (18.8) 68.4 (19.3) Immediately after episode 72.2 (18.6) 74.4 (17.7) 76.0 0.6 (2.9 to 4.1) 68.8 (20.5) 68.0 (21.3) 76.3 1.3 (2.7 to 5.2) Six months after episode 72.3 (18.4) 77.4 (14.9) 2.7 (6.5 to 1.2) 67.1 (21.1) 67.4 (21.1) 0.3 (4.2 to 4.8) a b c d Set at 74.0 for STAI, 1/3 of baseline s.d.s otherwise. Adjusted for baseline. Higher score indicates greater level of anxiety. Higher score indicates greater level of depression. Higher score indicates better health state. & 2006 Cancer Research UK British Journal of Cancer (2006) 95(4), 435 – 444 Clinical Studies Clinical Studies Genetic nurse counsellors can be an acceptable and cost-effective N Torrance et al Table 3 Health related quality of life, SF-36 scores : mean (s.d.) Grampian Wales b b Time Nurse counsellor Geneticist Difference (95% CI) Nurse counsellor Geneticist Difference (95% CI) Physical functioning Baseline 88.9 (18.6) 85.4 (21.8) 83.5 (21.8) 88.9 (18.3) FU1 88.2 (19.1) 88.6 (18.7) 0.5 (3.3 to 2.2) 84.9 (20.9) 88.0 (22.5) 0.3 (3.7 to 4.3) FU2 87.9 (18.5) 86.4 (21.3) 0.2 (2.3 to 2.7) 84.6 (21.3) 88.6 (20.1) 0.5 (5.5 to 4.5) Social functioning Baseline 84.0 (23.4) 84.6 (22.1) 77.7 (24.4) 79.8 (26.2) FU1 83.4 (23.9) 85.2 (21.2) 0.6 (5.7 to 4.5) 78.9 (24.8) 79.9 (25.4) 0.3 (4.9 to 5.4) FU2 84.7 (21.8) 87.2 (23.2) 1.0 (6.4 to 4.4) 78.0 (26.9) 80.1 (26.4) 1.0 (7.6 to 5.5) Role physical Baseline 87.1 (29.0) 86.1 (31.3) 81.6 (33.5) 84.9 (30.9) FU1 86.8 (30.2) 85.1 (32.6) 5.1 (1.6 to 11.8) 77.4 (37.7) 82.7 (33.4) 3.9 (11.4 to 3.6) FU2 86.0 (30.9) 87.1 (30.3) 1.3 (6.1 to 8.7) 77.5 (37.9) 74.0 (38.7) 5.5 (4.3 to 15.4) Vitality Baseline 58.6 (21.3) 58.5 (23.3) 53.6 (21.1) 54.9 (21.3) FU1 60.8 (21.7) 61.7 (19.4) 0.5 (3.7 to 4.7) 57.1 (22.3) 55.7 (20.3) 2.0 (2.3 to 6.3) FU2 61.6 (20.7) 63.9 (19.0) 1.4 (5.7 to 2.9) 55.3 (22.5) 58.3 (21.1) 1.8 (7.0 to 3.4) Bodily pain Baseline 76.3 (23.9) 76.6 (25.1) 72.3 (25.4) 75.5 (25.0) FU1 78.6 (24.7) 77.4 (24.4) 2.3 (2.4 to 7.1) 75.8 (24.6) 75.8 (26.0) 0.2 (5.3 to 4.9) FU2 78.2 (24.5) 76.1 (23.8) 1.7 (3.8 to 7.2) 74.9 (24.9) 75.2 (19.4) 0.2 (6.8 to 6.4) General health Baseline 73.5 (19.8) 73.4 (18.9) 66.0 (20.6) 71.2 (20.0) FU1 75.2 (20.7) 74.9 (18.4) 0.8 (2.5 to 4.0) 67.9 (21.4) 69.9 (20.7) 1.0 (2.4 to 4.3) FU2 75.0 (18.6) 73.7 (18.5) 1.7 (2.0 to 5.4) 68.6 (21.5) 72.5 (19.4) 0.03 (3.9 to 3.8) a b c Higher score indicates better health state (range 0 – 100). Adjusted for baseline. Larger than expected difference due to baseline imbalance of responders at FU1. Table 4 Respondents indicating ‘Strongly Agree/ Agree’ with statements on causes of breast cancer, n (%) Grampian Wales a a Nurse counsellor Geneticist P-value Nurse counsellor Geneticist P-value Stress is a major cause of breast cancer Baseline 68 (36) 40 (42) 68 (35) 34 (35) Immediately after episode 80 (47) 39 (48) 0.98 63 (38) 31 (37) 0.98 Six months after episode 71 (42) 31 (44) 0.83 55 (37) 26 (36) 0.91 One close relative with breast cancer always increases your risk considerably Baseline 157 (83) 77 (81) 165 (84) 80 (83) Immediately after episode 127 (73) 52 (63) 0.12 130 (77) 65 (77) 0.99 Six months after episode 123 (77) 56 (76) 0.97 122 (81) 58 (80) 0.88 Minor injury can cause breast cancer Baseline 37 (20) 20 (21) 62 (32) 23 (24) Immediately after episode 38 (22) 22 (27) 0.50 50 (30) 16 (19) 0.09 Six months after episode 47 (29) 17 (23) 0.39 50 (33) 18 (25) 0.24 a 2 From w test (Yates’ corrected – for 2 2 table). scale, for Grampian and Wales, respectively. The data suggested Patient knowledge ‘equivalence’ in these outcomes, in both trials, at both follow-up points, with the exception of the mental health score at the second Table 4 summarises the proportions of participants who agreed or follow-up point in the Grampian trial, which indicated that strongly agreed with each statement on a possible cause of breast ‘uncertain equivalence’. cancer. The results are generally similar between the two trial Table 3 summarises the SF36 scores for the other health status locations, and between intervention and control arms within each outcomes. Scores were generally high, the lowest being observed trial. Misunderstanding was greatest for the effect of having a first- for the vitality domain in both trials. On average, higher scores degree relative with breast cancer. No consistent, or statistically were observed in all domains in the Grampian trial compared with significant improvements in knowledge were observed for any of the Wales trial. the three notional risk factors. British Journal of Cancer (2006) 95(4), 435 – 444 & 2006 Cancer Research UK Genetic nurse counsellors can be an acceptable and cost-effective N Torrance et al & 2006 Cancer Research UK British Journal of Cancer (2006) 95(4), 435 – 444 Grampian FU1 (immediately after counselling episode) FU2 (6 months later) Given information wanted about breast cancer causes Given information wanted about breast cancer causes Given information wanted about breast cancer risks Given information wanted about breast cancer risks Given information wanted about breast cancer prevention Given information wanted about breast cancer prevention Given information wanted about genetic tests Given information wanted about genetic tests The doctor/nurse listened Concerns taken seriously Concerns taken seriously The doctor/nurse listened Satisfied with waiting time for first appointment Satisfied with waiting time for first appointment The consultation was helpful The consultation was helpful Satisfied overall Nurse counsellor 0 20406080 100 Satisfied overall Geneticist % Agree/strongly agree Nurse counsellor 0 20406080 100 Geneticist % Agree/strongly agree Wales FU1 (immediately after counselling episode) FU2 (6 months later) Given information wanted about breast cancer causes Given information wanted about breast cancer causes Given information wanted about breast cancer risks Given information wanted about breast cancer risks Given information wanted about breast cancer prevention Given information wanted about breast cancer prevention Given information wanted about genetic tests Given information wanted about genetic tests Concerns taken seriously The doctor/nurse listened The doctor/nurse listened Satisfied with waiting time for first appointment Satisfied with waiting time for first appointment Concerns taken seriously The consultation was helpful The consultation was helpful Satisfied overall Satisfied overall Nurse counsellor Nurse counsellor 0 20406080 100 0 20406080 100 Geneticist Geneticist % Agree/strongly agree % Agree/strongly agree Figure 2 Patient satisfaction with genetic counselling. Clinical Studies Clinical Studies Genetic nurse counsellors can be an acceptable and cost-effective N Torrance et al Patient satisfaction DISCUSSION High levels of patient satisfaction were observed in both trials. This study suggests that, for the initial episode of genetic Figure 2 shows the views of patients on the specific aspects of counselling for risk of breast cancer, nurse counsellors can provide satisfaction with services immediately following the genetic care that appears to be equivalent to that provided by clinical counselling episode and at the 6 month follow-up point. geneticists, in terms of patients’ psychosocial outcomes and satisfaction. The similar findings in the two separate trials support the generalisability of the findings. Acceptability to referring GPs Response rates to the baseline and follow-up surveys were high and we were able to assess the stability of the outcomes over a 6- In all, 74 and 87 GPs in Grampian and Wales, respectively, referred month period following the counselling episode. A high proportion at least one patient who was randomised to the nurse counsellor of eligible patients were recruited into both trials, suggesting that and who attended the genetic clinic. In all, 68 GPs in Grampian our study populations were representative of the target popula- (response rate 92%) and 75 GPs in Wales (response rate 86%) tions. The small but distinct baseline differences between the two responded to the acceptability survey. Grampian respondents, 60 sets of trial participants probably reflect real differences in the (88%) and 52 (69%) Wales respondents could not differentiate patient populations from which they were drawn. Participants in whether their patient had been seen by a nurse counsellor or a Wales had slightly higher anxiety levels overall than the Grampian clinical geneticist. Almost all respondents (Grampian 100%, Wales participants, but these were balanced between intervention and 98.7% (n¼ 74)) reported that they would be happy for future control arms within each setting. After genetic counselling, small referred patients to be seen at the clinic by the genetic nurse reductions in anxiety levels were seen in all groups. Even though counsellor. Overall satisfaction with the medical genetics service the participants’ baseline mean anxiety scores were slightly higher was high with 91% (62 out of 68) GPs in Grampian and 89% (67out than the norm for adult women (score of 35), they were of 75). In Wales reporting that they were ‘very satisfied/satisfied’ comparable to anxiety levels found in other studies that have with the service provided by the respective medical genetics used the STAI (Cull et al, 1998; Cull et al, 1999; Julian-Reynier services. et al, 1999; Brain et al, 2000). The between-trial differences support the decision to conduct two parallel trials rather than multicentre trial, where the data would have been pooled (Bowling, 1997). Economic evaluation The unit costs per counselling appointment Comparative data for SF-36 scores for these two general were similar for the clinical geneticist arms in both trials. Cost geographical populations (Garratt et al, 1993; Lyons et al, 1995) differences between intervention and control arms across the two also suggest that our trial participants were slightly more anxious locations were largely driven by staff costs. For Grampian, the than the underlying populations, but the between-trial differences marginal (additional) cost per single counselling appointment was were similar to the background between-population differences. It d17.98 higher for the control compared with the intervention arm; is also possible that these apparent differences reflect the different in Wales, the marginal cost per counselling appointment was healthcare referral processes in operation in the two locations. For d12.50 higher for the intervention than the control arm. example, in Wales, women who were considered to be at increased Table 5 presents the health service costs per counselling episode. familial risk at the breast screening clinic (Breast Test Wales) were For Grampian women randomised to the nurse counsellor, the first referred to a surgeon who, after review, made the referral for ‘counselling episode’ ranged from 1 to 4 appointments. Most genetic counselling. In Grampian, many women were referred women received one (149 out of 193, 77%) or two (38, 20%) more directly by GPs or other providers to the genetics clinic. It is appointments. In the control arm, 81 out of 96 (84%) participants possible that these differences in referral pathways contributed to received one, and 13 (13.5%) participants received two, appoint- ments. The mean cost per patient for the counselling episode was different anxiety levels in women by the time they received a d136.55 for the nurse counsellor arm and d148.30 for the clinical genetics clinic appointment, and also influenced their risk perceptions. Despite these apparent population differences, the geneticist arm, a difference of d11.54 (95% CI, d25.43, 1.94). In effectiveness data from the two trials were very similar. Wales all participants received one appointment only, hence the Few of the participants in these trials would have been eligible cost per episode and appointment were the same. for genetic testing, and this was not the focus of the study (which Overall, the cost per counselling episode for the nurse counsellor was the period in which initial assessments were made). In both was d3.55 lower in Grampian compared with Wales, and for the locations, women who were assessed as suitable for genetic testing geneticist arm was d20.70 higher in Grampian than in Wales. would have been referred to a consultant geneticist for follow-up Health service unit costs in both centres appeared sensitive to care. Our study end point was the point at which a woman learnt of grades of staff employed (nurses and doctors), level of supervision her risk status, and further follow-up or management arrange- of nurse specialists, and length of counselling appointments, but ments were made. We do not have access to data on how many of not to choice of discount rate or lifespan of equipment items. Table 5 Comparison of health service costs per counselling episode Grampian Wales Mean number of randomised Unit Mean total cost per Mean number of randomised Unit Mean total cost per a a Group appointments (range) cost patient (d) appointments (range) cost patient (d) Nurse counsellor 1.26 (1 – 4) 108.01 136.55 1 (1 – 1) 140.10 140.10 Geneticist 1.18 (1 – 3) 125.99 148.30 1 (1 – 1) 127.60 127.60 Difference in cost 11.54 +12.50 (nurse counsellor – geneticist) a b Unit cost of counselling appointment. (95% CI, d25.43, d1.94). British Journal of Cancer (2006) 95(4), 435 – 444 & 2006 Cancer Research UK Genetic nurse counsellors can be an acceptable and cost-effective N Torrance et al the participants actually went ahead with genetic testing, or their assessment service include the concept of working in liaison or actual test results. outreach settings (Emery et al, 1999; Fry et al, 1999). Two The primary outcome measure was anxiety, the reduction of which randomised controlled trials have reported on different models of is regarded as a key counselling objective (Shaw et al, 1999; Brain service delivery for genetic counselling for risk of breast cancer et al, 2000; Meiser and Halliday, 2002) and a number of evaluations (Brain et al, 2000; Fry et al, 2003). In the trial by Brain et al (2000), of genetic counselling for familial cancer have identified the pre- and the addition of specialist genetic assessment to the standard postcounselling assessment of generalised anxiety as a main outcome surgical consultation had no effect on patients’ psychological measure (Cull et al, 1998, 1999; Julian-Reynier et al, 1999; Brain et al, outcomes, risk perception or satisfaction, although knowledge of 2000; Kent et al, 2000; Bish et al, 2002; Bowen et al,2004).Evidence cancer genetics showed greater improvement. The other RCT from systematic reviews suggests that, overall, genetic counselling found community-based genetic nurse specialists to be generally has the effect of significantly reducing patients’ anxiety levels, at least comparable to the standard service (consultant geneticist), in in the short-term (Meiser and Halliday, 2002; Butow et al, 2003; terms of psychosocial outcomes and patient satisfaction, with the Braithwaite et al, 2004). The apriori equivalence limits for the additional benefit of lower staff and patient costs (Fry et al, 2003; primary outcome (STAI) were set at a very strict level, and in reality Holloway et al, 2004). they probably represent a smaller difference than would normally be There is now an emerging body of research evidence on the considered clinically significant between two clinicians considered relative effectiveness of nurses in specialist roles compared with equally competent; however, results showed that some outcomes doctors in both primary care (Horrocks et al, 2002), and secondary were considered ‘equivalent’ even at this strict level. care settings (Kinley et al, 2001; Sharples et al, 2002), with most Three common mistaken beliefs on the causes of breast cancer studies finding that nurses working to guidelines appear to provide were not influenced by counselling; the most persistent and care that is equal to that provided by doctors, with comparable erroneously held belief was in the influence of one close relative health outcomes. Previous randomised trials in secondary care with breast cancer on a person’s own risk. This may have reflected have found nurse practitioners to be either cost neutral (Kinley a general misconception, or reflect the participants’ own personal et al, 2001) or more expensive than doctors either because of salary risk perceptions. The knowledge we assessed was not specific to costs (Sakr et al, 2003) or greater resource use (Sharples et al, cancer genetics, and our findings are consistent with genetic 2002). In this study, we found that the relative costs of the nurse counselling delivered by either doctors or nurses being equally counsellors, compared with the doctor-led service, depended on effective (or ineffective) in educating patients about breast cancer the grade of medical staff whose time was being replaced by the risk. Braithwaite et al (2004) suggest that genetic counselling can nurse counsellor, and the extent of supervision required – of both be effective in improving knowledge related to breast cancer nurse counsellors and less experienced medical staff. Surprisingly, genetics, compared with no counselling or counselling delivered by the sensitivity analysis did not suggest that nurse counsellors of nongenetics specialists. However, the observed level of misunder- lower grades would be less costly than the nurses employed here, standing seen in this study suggests that more effective population on the basis that the cost saving in lower salaries would be offset by interventions are required to improve general knowledge of breast the greater need for consultant supervision for nurses at lower cancer risks. grades. Patient satisfaction with information provided, staff attitudes The results of this study add to the emerging body of evidence and the overall clinic procedures were high overall in both trials supporting the effectiveness, and possible cost-effectiveness, of irrespective of randomised group. Similar levels of patient nurse counsellors working under the supervision of consultant satisfaction have been reported in other trials of genetic geneticists, and should be taken into account by decision-makers counselling services for risk of breast cancer (Brain et al, 2000; planning and evaluating genetics health services. Hopwood et al, 2004; Holloway et al, 2004). In addition, the acceptability of the genetic nurses counsellors was high among referring GPs in both trial settings. However, it is possible that the ACKNOWLEDGEMENTS difference in service model (i.e. appointment of the nurse counsellors) may be imperceptible at primary care level as We thank Shelley Farrar, Richard Gordon, Tracy Ibbotson and individual GPs referred only one or two patients during the trial. Leslie Walker who were involved in the original grant application The lack of well-designed and rigorously conducted randomised and study design. Thanks go to the genetic nurse counsellors who controlled trials, with reporting to CONSORT standards, in the were involved in the study, and to all the NHS clinical and field of service delivery for genetic counselling for familial cancer administrative staff in both clinical genetics centres. We are has been observed in a number of recent publications (Butow et al, grateful to the patients who participated in both trials. This study 2003; Braithwaite et al, 2004; Hopwood et al, 2004; NICE, 2004). was funded by the NHS Health Technology Assessment pro- Alternative models for how nurse counsellors might provide a risk gramme. REFERENCES Brain K, Gray J, Norman P, France E, Anglim C, Barton G, Parsons E, Bowen DJ, Burke W, McTiernan A, Yasui Y, Andersen MR (2004) Breast Clarke A, Sweetland H, Tischkowitz M, Myring J, Stansfield K, Webster cancer risk counseling improves women’s functioning. Patient Edu D, Gower-Thomas K, Daoud R, Gateley C, Monypenny I, Singhal H, Couns 53: 79– 86 Branston L, Sampson J, Roberts E, Newcombe R, Cohen D, Rogers C, Bowling A (1997) Research Methods in Health: investigating health and Mansel R, Harper P (2000) Randomized trial of a specialist genetic health services. Buckingham: Open University Press assessment service for familial breast cancer. J Natl Cancer Inst 92(16): Butow PN, Lobb EA, Meiser B, Barratt A, Tucker KM (2003) Psychological 1345– 1351 outcomes and risk perception after genetic testing and counselling in Braithwaite D, Emery J, Walter F, Prevost AT, Sutton S (2004) Psychological breast cancer: a systematic review. Med J Aust 78: 77– 81 impact of genetic counselling for familial cancer: a systematic review and Claus EB, Risch N, Thompson WD (1994) Autosomal dominant inheritance meta-analysis. J Natl Cancer Inst 96(20): 122– 133 of early onset breast caner. Cancer 73: 643–651 Bish A, Sutton S, Jacobs C, Levene S, Ramirez A, Hodgson S (2002) Changes Cull A, Anderson EDC, Campbell S, Mackay J, Smyth E, Steel CM (1999) in psychological distress after cancer genetic counselling: a comparison The impact of genetic counselling about breast cancer risk on women’s of affected and unaffected women. Br J Cancer 86: 43–50 risk perceptions and levels of distress. Br J Cancer 79(3/4): 501 –508 & 2006 Cancer Research UK British Journal of Cancer (2006) 95(4), 435 – 444 Clinical Studies Clinical Studies Genetic nurse counsellors can be an acceptable and cost-effective N Torrance et al Cull A, Miller H, Porterfield T, Mackay J, Anderson ED, Steel CM, Elton RA gical morbidity and health beliefs in women attending for genetic (1998) The use of videotaped information in cancer genetic counselling: counselling. Br J Cancer 74: 482–487 A randomised evaluation study. Br J Cancer 77(5): 830–837 Lyons RA, Fielder H, Littlepage BN (1995) Measuring health status with the Emery J, Watson E, Rose P, Andermann A (1999) A systematic review of the SF-36: the need for regional norms. J Public Health Med 17: 46–50 literature exploring the role of primary care in genetic services. Fam Marteau TM, Bekker HB (1992) The development of a six-item short-form Pract 16: 426–445 of the state scale of the Spielberger State-Trait Anxiety Inventory (STAI). Fry A, Campbell H, Gudmunsdottir H, Rush R, Porteous M, Gorman D, Cull Br J Clin Psychol 31: 301–306 A (1999) GPs’ views on their role in cancer genetics services and current Marteau TM, Walker P, Giles J, Smail M (1990) Anxieties in women practice. Fam Pract 16: 468–474 undergoing colposcopy. Br J Obstet Gynaecol 97: 859 –861 Fry A, Cull A, Appleton S, Rush R, Holloway S, Gorman D, Cetnarskyj R, Meiser B, Halliday JL (2002) What is the impact of genetic counselling in Thomas R, Campbell J, Anderson E, Steel M, Porteous M, Campbell H women at increased risk of developing hereditary breast cancer? A meta- (2003) A randomised trial of breast cancer genetics services in South East analytic review. Soc Sci Med 54: 1463–1470 Scotland: psychological impact. Br J Cancer 89: 653– 659 Miedzybrodzka ZH, Hall MH, Mollison J, Templeton A, Russell IT, Dean JC, Garratt AM, Ruta DA, Abdalla MI, Buckingham JK, Russell IT (1993) The Kelly KF, Marteau TM, Haites NE (1995) Antenatal screening for carriers SF36 health survey questionnaire: an outcome measure suitable for of cystic fibrosis: randomised trial of stepwise v couple screening. BMJ routine use within the NHS? BMJ 306: 1440 –1444 310: 353–357 Holloway S, Porteous M, Cetnarskyj R, Anderson E, Rush R, Fry A, Gorman National Institute for Clinical Excellence (2004) Clinical Guidelines and D, Steel M, Campbell H (2004) Patient satisfaction with two different Evidence Review for the Classification and Care of Women at Risk of models of cancer genetic services in south-east Scotland. Br J Cancer 90: Familial Breast Cancer. London: National Collaborating Centre for 582–589 Primary Care/University of Sheffield, National Institute for Clinical Hopwood P, Wonderling D, Watson M, Cull A, Douglas F, Cole T, Eccles D, Excellence (www.nice.org.uk) Gray J, Murday V, Steel M, Burn J, McPherson K (2004) A randomised Office for National Statistics (2005) http://www.statistics.gov.uk/cci/ comparison of UK genetic risk counselling services for familial cancer: nugget.asp?id¼ 575 [accessed 08/12/2005] psychosocial outcomes. Br J Cancer 91(5): 884 –892 Pocock SJ (1983) Clinical trials: a practical approach. Chichester: Wiley Horrocks S, Anderson E, Salisbury C (2002) Systematic review of whether Sakr M, Kendall R, Angus J, Saunders A, Nicholl J, Wardrope J (2003) nurse practitioners working in primary care can provide equivalent care Emergency nurse practitioners: a three part study in clinical and cost to doctors. BMJ 324: 819 –823 effectiveness. Emerg Med J 20: 158– 163 Jones B, Jarvis P, Lewis JA, Ebbutt AF (1996) Trials to assess equivalence: Sharples LD, Edmunds J, Bilton D, Hollingworth W, Caine N, Keogan M, the importance of rigorous methods. BMJ 313: 36–39 Exley A (2002) A randomised controlled crossover trial of nurse Julian-Reynier C, Eisinger F, Chabal F, Aurran Y, Bignon YI, Machelard- practitioner versus doctor led outpatient care in a bronchiectasis clinic. Roumagnac M, Maugard C, Nogues C, Vennin P, Sobol H (1999) Cancer Thorax 57: 661–666 genetic consultation and anxiety in healthy consultees. Psychol Health 14: Shaw C, Abrams K, Marteau TM (1999) Psychological impact of predicting 379–390 individuals’ risks of illness: a systematic review. Soc Sci Med 49: 1571– Kent G, Howie H, Fletcher M, Newbury-Ecob R, Hosie K (2000) The 1598 relationship between perceived risk, thought intrusiveness and emotional Shiloh S, Avdor O, Goodman RM (1990) Satisfaction with genetic well-being in women receiving counselling for breast cancer risk in a counseling: dimensions and measurement. Am J Med Genet 37: 522–529 family history clinic. Br J Health Psychol 5: 15– 26 Skirton H, Barnes C, Guilbert P, Kershaw A, Kerzin-Storrar L, Patch C, Kinley H, Czoski-Murray C, George S, McCabe C, Primrose J, Reilly C, Curtis G, Walford-Moor J (1998) Recommendations for education and Wood R, Nicolson P, Healy C, Read S, Norman J, Janke E, Alhameed H, training of genetic nurses and counsellors in the United Kingdom. J Med Fernandez N, Thomas E (2001) Extended scope of nursing practice: a Genet 35: 410–412 multicentre randomised controlled trial of appropriately trained nurses Ware JE, Kosinski M, Keller SD (1994) SF36 Physical and mental and pre-registration house officers in pre-operative assessment in component summary measures: a users manual. The Health Institute: elective general surgery. Health Technol Assess 5(20): 1–87 Boston Kinmonth AL, Reinhard J, Bobrow M, Pauker S (1998) The new genetics. Watson M, Lloyd S, Davidson J, Meyer L, Eeles R, Ebbs S, Murday V (1999) Implications for clinical services in Britain and the United States. BMJ The impact of genetic counselling on risk perception and mental health 316: 67–70 in women with a family history of breast cancer. Br J Cancer 79(5 –6): Lerman C, Daly M, Sands C, Balshem A, Lustbader E, Heggan T, Goldstein 868–874 L, James J, Engstrom P (1993) Mammography adherence and Wonderling D, Hopwood P, Cull A, Douglas F, Watson M, Burn J, psychological distress among women at risk for breast cancer. J Natl McPherson K (2001) A descriptive study of UK cancer genetics services: Cancer Inst 85: 1074– 1080 an emerging clinical response to the new genetics. Br J Cancer 85: 166–170 Lloyd S, Watson M, Waites B, Meyer L, Eeles R, Ebbs S, Tyle A (1996) Zigmond AS, Snaith RP (1983) The hospital anxiety and depression scale. Familial breast cancer: a controlled study of risk perception, psycholo- Acta Psychiatr Scand 67: 361–370 British Journal of Cancer (2006) 95(4), 435 – 444 & 2006 Cancer Research UK

Journal

British Journal of CancerSpringer Journals

Published: Jul 11, 2006

There are no references for this article.