Office of Population Health Genomics, Department of Health, Western Australia; Centre for Population Health Research, Curtin Health Innovation Research Institute (CHIRI), Curtin University of Technology, Western Australia, and School of Population Health, University of Western Australia
Office of Population Health Genomics, Department of Health, Western Australia; Centre for Population Health Research, Curtin Health Innovation Research Institute (CHIRI), Curtin University of Technology; School of Pathology and Laboratory Medicine, University of Western Australia and School of Women's and Infants’ Health, University of Western Australia
Correspondence to: Ms Susannah Maxwell, Office of Population Health Genomics, Third floor C Block, 189 Royal Street, East Perth, WA 6004; e-mail: firstname.lastname@example.org
Objective: Professional guidelines define the risk categorisation of patients for a genetic predisposition to cancer based on family history. These guidelines inform the appropriate referral of patients to specialist familial cancer services. Our study aimed to determine the quality of referral letters from general practitioners and specialists to genetic services for breast, ovarian and colorectal cancers, and their compliance with relevant professional guidelines.
Methods: A retrospective review of the referral letters and patient files of 241 consecutive patients referred between June and October 2008.
Results: Sufficient information to make a risk assessment was provided in 71% of referrals. Of these, 89% were compliant with guidelines. Genetic counsellors collected further information on 167 of the 241 referred patients and of these 83% were appropriate for referral according to guidelines.
Conclusions and implications: Overall, referrals to familial cancer genetic services complied with professional referral guidelines. The majority of referrals were high quality, and with additional information, most patients were shown to be appropriate for review in a familial cancer clinic. Despite this, a better understanding of the reasons for non compliant referrals, and appropriate targeted education and resources is recommended to improve referral quality and compliance.
Genetic medicine has expanded rapidly over the past decade and there are now more than 1,550 diseases for which genetic tests are available,1 the majority of which are for single gene disorders. Demand on genetic services has risen accordingly. Although most tests are for rare diseases, tests to identify inherited risk for common diseases such as breast, ovarian and colorectal cancer, well established as part of clinical care,2 now account for around half the workload of our clinical genetics services. Early identification of individuals with an inherited predisposition to cancer and appropriate management has been shown to be a cost-effective strategy for reducing cancer related morbidity and mortality.3,4 However, cost-effectiveness relies on prioritising genetic tests to those triaged on their family history as having a certain likelihood of segregating a high risk cancer susceptibility gene mutation.5
It is estimated that 5–10% of common cancers are attributable to inherited risk,6–7 and the optimal referral rate for individuals with cancer to genetic services is considered to be as low as 6%.8 Identifying patients at increased risk of a genetic predisposition to cancer is an important initial step in the management of patients with a family history. Australian guidelines developed by the National Health and Medical Research Council (NHMRC) in 1999, which defined the categorisation of patients as low, moderate and high risk based on family history information inform the appropriate referral of patients to specialist cancer services.7 These guidelines are similar to those produced by the American College of Obstetricians and Gynecologists9 (2009) for genetic testing of women with cancer and the US Preventive Services Task Force (USPSTF) (2005) guidelines for asymptomatic breast and ovarian cancer.10
Other Australian guidelines produced by the National Breast Ovarian Cancer Centre (NBOCC) (2006)11 and the Australian Cancer Network (ACN) (2006)9 include advice about familial aspects of breast cancer and ovarian cancer and familial aspects of bowel cancer respectively. According to these guidelines, referral to clinical genetic services is recommended if the individual is at moderately increased or potentially high risk for breast/ovarian cancer, or high risk for bowel cancer (Table I). For individuals at low risk, referral to specialist familial cancer services is considered unnecessary and appropriate management can be provided through the individual's treating practitioner.
Increasing public awareness of inherited cancers and predictive genetic testing, together with the limited number of genetic counsellors, has placed greater attention on the role of the primary care physician in assessment and referral to familial cancer services.12 However, in a recent study of cancer genetic referral patterns in primary care, Vig and colleagues13 (2009) reported less than 17% of primary care physicians affiliated with a large US health insurance carrier used professional referral guidelines.
In Western Australia, the Familial Cancer Program of Genetic Services WA (GSWA) provides a full range of diagnostic and counselling services for individuals at increased risk of familial breast, ovarian and colorectal cancers. Individuals are referred through general practitioners, other medical specialists, indirectly through other family members who have attended GSWA as patients or through a process of self-referral. An understanding of referral patterns and characteristics is essential to guide the development of efficient services. This study examined consecutive referrals for breast, ovarian and colorectal cancers from general practitioners and specialists (surgeons and physicians) to Genetic Services Western Australia (GSWA) to determine compliance with the relevant professional guidelines.9,11
A retrospective audit of general practitioner and specialist referrals to GSWA was conducted for the period June 2008 to October 2008 (five months). The audit of referrals was restricted to those from health professionals for potential hereditary breast and ovarian cancer and potential hereditary non polyposis colorectal cancer (HNPCC or Lynch syndrome).
Referred clients were identified via a search of the GSWA database (KinTrak™) using the following terms: colorectal cancer, HNPCC, breast cancer, ovarian cancer, endometrial cancer, colon cancer and bowel cancer. Client records were also reviewed to confirm they were referred by a general practitioner or specialist. Self-referred or family-referred clients, and referrals for predictive testing for any known mutation, were excluded.
Each referral letter and patient file was examined by a genetic counsellor (KH) and recorded in a database developed in Microsoft Access. We identified if the referral was from a general practitioner or a specialist (the type of specialist was not recorded). We collected information on the reason for referral, the presence or absence of a relevant comprehensive family history and evidence of risk stratification in the referral. The patient file provided additional family history information collected by the genetic counsellors. All patients referred to genetic counselling receive an offer of appointment letter and/or telephone contact.
Each referral was categorised according to quality and compliance with guidelines before and after an appointment with GSWA, as shown in Figure 1.
Family history information required for risk assessment includes age of diagnosis, primary cancer site, sex and relationship to the index case as shown in Table 1.9,11 Referrals were categorised as high or low quality depending on the content of the referral letter. A high quality referral contained sufficient family history information to determine the patient's risk using professional guidelines while a low quality referral contained insufficient family history information for risk assessment.
Table 1. Breast/ovarian and colorectal cancer guidelines: Risk categories for which referral to genetic services is appropriate.
Note: a Without the additional features of the high risk breast/ovarian cancer group Adapted from National Breast Ovarian Cancer Centre. Advice about familial aspects of breast cancer and epithelial ovarian cancer: a guide for health professionals. 2006, Australian Cancer Network. Familial Aspects of Bowel Cancer: A guide for health professionals. 2006.
Moderately increased risk 1.One 1° relative with breast cancer before the age of 50a 2.Two 1° relatives on the same side of the family, diagnosed with breast cancera 3.Two 2° relatives, on the same side of the family, diagnosed with breast cancer, at least one before the age of 50a
Potentially high risk 1. Two 1° or 2° relatives on one side of the family diagnosed with breast or ovarian cancer plus one or more of the following features on the same side of the family: • Additional relative(s) with breast or ovarian cancer • Breast cancer diagnosed before the age of 40 • Bilateral breast cancer • Breast and ovarian cancer in the same woman • Ashkenazi Jewish ancestry • Breast cancer in a male relative 2. One 1° or 2° relative diagnosed with breast cancer at age 45 or younger plus another 1° or 2° relative on the same side of the family with sarcoma (bone/soft tissue) at age 45 or younger
Potentially high risk 1. Three or more 1° relatives or a combination of 1° or 2° relatives on the same side of the family diagnosed with bowel cancer 2. Two or more 1° or 2° relatives on the same side of the family diagnosed with bowel cancer, plus any of the following high risk features: • Bowel cancer before the age of 50 • A family member who has/had an HNPCC related cancer (endometrial, ovarian, stomach, small bowel, renal pelvis or ureter, biliary tract, brain cancer) 3. At least 1° or 2° relative with a large number of adenomas throughout the large bowel (suspected FAP)
Referral compliance with guidelines
High quality referrals were examined for compliance with relevant guidelines (Table 1). Referral is compliant with guidelines if the patient is at a moderately increased risk or potentially high risk for breast cancer, and potentially high risk for colorectal cancer. In the case where the patient being referred to GSWA has had a personal diagnosis of cancer, in the absence of guidelines for the referral of those with a diagnosis, the NBOCC guidelines for asymptomatic patients were used with the patient included in the family history (i.e. a 30 year old with breast cancer, would have been included as a first degree relative under the age of 50 with breast cancer).
Post hoc patient assessment
Genetic counsellors verified and collected further information on each referred patient to determine if it was appropriate for the patient to be referred to genetic services. This review included the validation of family history information. A patient was said to be appropriate for referral if, with the additional information collected by the genetic counsellor, they were considered to be at increased risk, as per the relevant guideline. After reviews, patients with high quality referrals that complied with guidelines may be shown to be inappropriate patients for referral, for example where the review of the initial information provided about the patient is found to be inaccurate. Further information collected by a genetic counsellor about low-quality referrals may also show that it is appropriate for the patient to be referred to genetic services according to guidelines. If no additional information was collected by genetic counsellors (i.e. there had been no appointment or no patient contact with genetic services) the patient was classified as unknown.
Statistical analysis was conducted using SPSS for Windows (v 17).
The source, type and reason for referral, and quality and compliance of the referral with guidelines are shown in Table 2. Of the 241 referrals, genetic counsellors were able to collect further information on 167 patients. The 74 patients who did not attend an appointment or could not be contacted were classified as ‘unknown’. Cases were equally as likely to be classified as ‘unknown’ in both the high and low quality referral categories and for specialists and general practitioners (Fisher's exact test p>0.05). Patients referred due to a cancer diagnosis were less likely to be classified as ‘unknown’ (p=0.02). In general, the failure to have an appointment was not due to a lack of resources at genetic services, but rather because the patient had not responded to the invitation to attend the Familial Cancer Program. Specialists referred a significantly higher proportion of patients with cancer than general practitioners (p<0.001). There was no significant difference between general practitioner and specialist referrals with regard to quality, compliance or appropriateness of referrals according to guidelines (p>0.05).
Table 2. Characteristics of referrals to Genetic Services WA.
GP % (n)
Specialist % (n)
Total % (n)
a) High quality referrals only, compliance of referral with guidelines (moderately increased risk or potentially high risk for breast cancer, potentially high risk for colorectal cancer).
b) Referral warranted according to guidelines after the collection of further information by genetic services. Excludes ‘unknown’.
Reason for referral
Relevant family history
Personal cancer diagnosis
Specialist clinic request
Patient request to doctor
With limited health resources, and increasing demand on clinical genetic programs, audits of quality and compliance with referral guidelines are vital to facilitate demand management and delivery of cost-efficient services. Referral of low risk patients increases the burden on genetic services and may potentially limit access to those at higher risk.14 It also has important psychosocial implications as referral of low risk individuals may cause unnecessary anxiety and fear in patients that do not require referral.15 However, Brain and colleagues16 suggest that for some patients for whom referral would not be appropriate according to the guidelines, but who have a heightened perceived risk, referral and consultation may provide a benefit by reducing anxiety.
With regard to quality of referrals, the provision of sufficient information (i.e high-quality referral) enables efficient patient triage. Referrals that contain inadequate information require initial contact with the referring practitioner to request more information, before the nature and urgency of the referral can be assessed. This additional step has the potential to delay identification and access for high risk patients.14 Among those 74 patients who did not attend their appointments, the majority (70%) did not have a formal diagnosis of cancer and were likely to have been referred for predictive genetic testing due to family history. Their non-attendance most likely reflects a personal decision not to seek further information about their risk of developing cancer at the time.
In our study, we found the majority of referrals to genetic services (89%) are compliant with Australian referral guidelines (Figure 1). This is slightly higher than the results of a 2007 UK study where the authors reported 79% of referrals to be compliant with guidelines.17 Another study in the UK in 2001 found only 61% of general practitioner referrals were compliant with referral guidelines.18 This variation may reflect regional differences or a growing awareness of referral guidelines for familial cancer over the past decade.
For the small number of patients (11%) whose referrals included family history information but did not comply with the guidelines, the reason for referral is unclear. While these referrals may reflect the failure to use or interpret the guidelines accurately, previous studies have shown some practitioners intentionally refer low-risk patients as defined by the guidelines. Reasons for this may include a perceived lack of harm in referring, maintaining a good relationship with the patient and an expectation that genetic services will appropriately counsel the patient about genetic testing.14,19
While most referrals were high quality, there was still a large proportion (29%) which failed to provide sufficient family history information. This is consistent with a UK study, which found one- quarter of referrals to a familial cancer clinic provided insufficient clinical information.18 The low quality referrals in our study may not necessarily reflect a lack of use of clinical guidelines, as further review showed that once additional information was collected, three-quarters of these patients were considered appropriate according to the guidelines (excluding ‘unknown’). While the provision of family history information is requested, current guidelines do not define precisely what details should be included in referrals. Overall, for the majority (83%) of patients for whom additional information was collected referral to genetic services was appropriate, according to the guidelines, slightly higher than the 73% and 75% reported in two UK studies.17,20
As the majority of the referrals in our study are of a high standard the most appropriate strategy may be to target education to referring practitioners on a case-by-case basis in response to a low-quality or non-compliant referral. Genetic education for health professionals and computer decision support aids have been shown to improve the identification of those individuals at high risk and improve the compliance of referrals with relevant guidelines.17,21 The provision of a specific familial cancer referral pad (electronic or hard copy) that refers to the guidelines and provides prompts for information may act to both inform and direct referring practitioners.
The development of targeted individualised interventions for practitioners providing high-quality, yet non-compliant referrals will require an understanding of why this is occurring. Our study did not consider those patients who may have benefited from referral but for whom a referral was not made. Research into the extent to which practitioners fail to refer and why, will help to complete the picture as to what education and resources are required to improve referrals and access to genetic services in Western Australia.
Overall, in Western Australia, referrals to familial cancer genetic services by referring practitioners complied with professional referral guidelines. The majority of referrals were high-quality and, with additional information, most patients were shown to be appropriate for review in the familial cancer service according to guidelines. Despite this, there remains room for improvement. A better understanding of the reasons for non-compliant referrals and the attitudes of local referring practitioners, and appropriate targeted education and resources are recommended to improve referral quality and compliance. Additionally, of concern is the large proportion of referred patients (31%) who had not accessed genetic counselling services at the time of review. Exploration of the sociodemographic characteristics of those who did and did not attend was beyond the scope of our study, but warrants further investigation.
The authors thank Leslie Colvin James and Helen Mountain, (Genetic Counsellors at Genetic Services of Western Australia) for advice on methodology and referral processes. Ethical approval for the study was granted by the Human Research Ethics Committee of the Department of Health Western Australia.