H Daly, C Collins
Irish College of General Practitioners, 4-5 Lincoln Place, Dublin 2

Abstract
This is the first needs assessment of general practitioners (GPs) in relation to early detection of cancer in the Republic of Ireland. Data was collected using focus groups with 47 GPs and a national survey of 929 GPs. Overall,74.6% of survey respondents had >10 years experience in general practice and 22.7% were single handed. Distance from the practice to the nearest hospital to which they could refer patients for assessment of suspected cancer was <5 miles for 53.3% of GPs. The principle barriers identified were not confined to early diagnosis but apply to the diagnosis of cancer at any stage. These included delayed patient presentation, lack of direct GP access to investigations, difficulty with hospital referral, lack of clear recommendations for screening, poor communication with hospital services, inequitable access to hospital services and a need for further education and clinical practice guidelines. The barriers identified have serious implications for early detection of cancer in primary care and are remedial.

Introduction
Although cancer is a major cause of morbidity and mortality in the Republic of Ireland detection of a patient with cancer is an uncommon event in primary care. There are an average 20,523 new cases of cancer registered per annum¹ and an estimated 2,477 GPs in Ireland.² An individual GP can expect to see approximately eight patients with a new cancer per year. Nonetheless, the vast majority of patients with cancer present initially to their GP.³ Differentiation between patients whose symptoms may be due to cancer and the much larger number with similar symptoms due to other causes represents a considerable challenge for a GP. The Irish Cancer Society (ICS) and Irish College of General Practitioners (ICGP) collaborated in a needs’ assessment of GPs to identify the barriers experienced by GPs to early detection of cancer.

Methods
Early detection was defined as prompt and effective referral of early symptomatic patients and identification and screening of‘high risk asymptomatic individuals’. Total population screening was excluded. Five focus groups were conducted with 47 GPs selected through the ICGP continuing medical education (CME) network on the basis of geographical convenience and willingness to participate. Sessions lasted one and a half hours and transcripts were analysed by hand using a content analysis approach. A questionnaire was also posted to all GPs in the ICGP membership database, excluding retired GPs and GP trainees; 956 completed questionnaires were returned, a response rate of 46.7%.The analysis was conducted using SPSS v14 on 929 replies received by the data lock date.

Results
Qualitative research: The principle barriers identified were delayed patient presentation, lack of direct GP access to radiological and endoscopic investigations, difficulty with referral of patients to hospital services, lack of clear recommendations for cancer screening, poor communication with hospital services and inequitable access to hospital services for patients who cannot afford to pay privately. These were not confined to early diagnosis but applied to cancer diagnosis at any stage.

Quantitative research: The extent to which the above barriers applied to GPs nationwide was determined via postal survey. Of the analysed sample of GP respondents, 554 (60.5%) were male, 681 (74.6%) had >10 years experience in general practice, 196 (22.7%) were single handed and 720 (79.3%) had a practice nurse. Responses were received from all counties with 204 (23.8%) from Dublin. Distance from the practice to the nearest hospital to which they could refer patients for assessment of suspected cancer was <5 miles for 486 (53.3%) GPs, 5-25 miles for 317 (34.8%) and >25 miles for 108 (11.9%) GPs.

Overall, 767 (85.4%) GPs reported that men, and 561 (62.9%) that the elderly, are more likely to present late with symptomatic cancer.

The greatest barrier reported was lack of direct GP access to hospital investigative services resulting in delayed diagnosis due to long outpatient (OP) waiting times. GPs reported greater direct access to all investigations for private patients with the greatest discrepancy for CT scanning and mammography (Table 1).

The majority of GPs reported longer waiting times at the local public hospital for public compared to private patients (Table 2). Significant discrepancies were also reported in waiting time for investigation results with 20-30% more GPs reporting receipt within two weeks for private compared to public patients.

With the exception of symptomatic breast services, which were considered excellent and superior to those for other cancers, GPs reported difficulty in accessing OP services particularly for urgent appointments and many patients may have to wait for up to two months (Table 3). They consider this unacceptable and feel there should be agreed timelines for urgent and routine referrals.

Almost all GPs (n=915; 98.5%) considered a patient with obvious or suspected cancer with acute symptoms requiring treatment e.g. obstruction, with a clinically obvious (n=917; 98.7%) or suspected (n=879; 94.6%) cancer to be urgent whilst 663 (71.4%) considered a patient with non-specific symptoms and suspected cancer to be so. Only one-third reported that urgent patients with acute symptoms requiring treatment received an appointment on the day of referral (Table 3). Overall, one -fifth reported that urgent patients with acute symptoms requiring treatment wait longer than two weeks for an appointment.

Almost half of GPs (n=420; 48.2%) reported having a rapid access facility at their local hospital for ‘urgent’ referrals only, 316 (41.7%) for patients with clinical features of a site-specific cancer, 229 (27.0%) for any patient with suspected cancer and only 113 (14%) for patients with suspected cancer with non-specific symptoms.

Most GPs (80-95%) reported screening both General Medical Services (GMS) and private patients by manual breast examination, mammography, colonoscopy, digital rectal examination (DRE), prostate specific antigen (PSA) testing, cervical cytology and urinalysis; less than half with faecal occult blood (FOB) testing, complete body skin examination and only one-quarter screening for oral cancer. They reported offering screening equally to GMS and private patients with over 90% offering it to patients attending with unrelated complaints, 40- 50% at ‘well man or woman’ clinics and one-quarter by systematic call-up. The principle barriers to such screening reported were lack of clear recommendations for cancer screening and that screening is not funded by the GMS.

GPs reported poor communication with hospital services and considered this a major barrier to early detection (Table 4). However, only 615 (68.4%) GPs reported always indicating if a referral is urgent and only 403 (44.3%) if cancer is suspected on their referral letter to the hospital. Only 106 (11.6%) GPs reported being routinely informed by the ‘hospital’ of the date of their patient’s appointment.

The majority of GPs (n=771; 83.4%) considered the ability to pay privately always or usually affects access to referral services.

In order to improve early detection, 660 (71.0%) GPs reported they need rapid access clinics for assessment of patients with suspected cancer, 568 (61.1%) shorter waiting lists for urgent patients and 510 (54.9%) equal access to investigation and referral for GMS and private patients at their local public hospital(s). The factors which GPs reported would assist them most in early detection were agreed criteria for screening ‘high risk individuals’ (n=499; 53.7%), agreed referral criteria for suspected cancer (n=490; 52.7%), a ‘ring fenced’ budget for community diagnostic services (n=439; 47.3%), increased public awareness of early cancer symptoms (n=412; 44.4%) and earlier patient presentation to the GP (n=403; 43.4%).

GPs reported a need for further education and clinical practice guidelines on identification of ‘high-risk individuals’, ‘targeted’ cancer screening for ‘high risk individuals’, appropriate investigative pathways for suspected cancer, referral criteria for suspected cancer and identification of early stage cancer.

Discussion
The GP and practice profile are similar to those reported elsewhere.² A literature review revealed a lack of research on barriers to early detection of cancer in primary care.⁴ The principal barriers identified in this study have serious implications, are similar to those reported by Scottish GPs⁵ and are remedial.

Delay in patient presentation is a critical barrier to early detection of cancer over which GPs have little influence. There should be a national approach to public education on early symptoms of common cancers and benefits of early detection and treatment, screening and prevention led by the Department of Health and Children (DOH&C) and the Health Service Executive (HSE) which should be coordinated with the ICGP so that GPs are prepared for increased queries and can respond in a uniform manner. As the number of patients with cancer is expected to double over the next 15 years due to ageing of our population, the needs of the elderly should be specifically addressed.⁶

In general, patients present to their GP with symptoms which may be due to cancer among many other conditions. Initial investigation must be undertaken by GPs prior to referral to sitespecific cancer services and requires direct and prompt GP access to investigative services.The reported difficulties in access to investigative services, lack of rapid access referral facilities and inequity of access all mitigate against early detection.

The high proportion of GPs without direct access to barium x-ray, ultrasound, CT scanning or GI endoscopy for public patients, together with long waiting times for these, has serious implications for the diagnosis of gastrointestinal cancer which includes many of the fatal cancers common in Ireland and may explain the reported lower rate of investigation prior to referral in the Republic of Ireland⁷ compared to Scotland.⁵ The discrepancy in access for CT scanning and mammography between public and private patients cannot be explained by concerns of inappropriate radiological investigation as the same GPs have access for private but not public patients. Our findings are consistent with those of a recent Irish survey which found GPs have poorer direct access to radiology services than in the 1980-1990s² and justifies their request for ‘ringed fenced’ community diagnostic services. The low level of access to endoscopy for all patients may reflect inadequate endoscopic resources. The reported greater likelihood of longer waiting times for public patients for investigations at public hospitals suggests private patients may receive preferential treatment in public facilities.

The reported difficulties in obtaining an urgent OP appointment for a patient with suspected cancer at a public hospital are unacceptable and justify GPs proposal for agreed timelines and rapid access clinics. Symptomatic breast services should not be the exception but the standard to which all OP cancer services should aspire. The needs of patients with non-specific symptoms only, who may represent the majority of patients, need to be addressed.

Experienced GPs felt they have a high index of suspicion for cancer. They felt targeted screening with appropriate funding and back up facilities would be necessary to increase their level of early detection.

The reported communication difficulties between public hospitals and GPs are an unnecessary barrier to (early) diagnosis of cancer. Communication could be readily improved by provision of a directory detailing the investigative and referral services available for suspected cancer, the recommended mode of communication, agreed referral criteria and clear referral arrangements for urgent and non-urgent patients together with automatic notification of the date of patients’ appointments. GP Referral guidelines for suspected cancer already exist and could be readily adapted for national use or local use.⁸ However, first ‘urgent’ must be defined and agreed centrally.

In conclusion, early detection and preliminary investigation of most patients with cancer occurs in general practice which is not adequately resourced for this. The reported lack, and inequity, of access and long waiting times for investigations and referral together with the high proportion of GPs who reported delaying investigation and/or referral due to difficulty in accessing services have serious implications for the (early) diagnosis of cancer. The study highlighted many serious but remedial barriers which must be addressed to improve early cancer detection. Many of these are envisaged in the Strategy for Cancer Control in Ireland which recommends they be implemented by the HSE.6 However, the emphasis remains on hospitals and more attention to community needs including designated diagnostic facilities and expansion of rapid access facilities is required.

References

1. National Cancer Registry Ireland. Cancer in Ireland, 1994-2001: Incidence, Mortality and Treatment. Cork: National Cancer Registry Board, 2005.
   http://www.ncri.ie/pubs/pubfiles/report2005_2.pdf&#160;(Accesse d 2 Aug 2007).
2. O’Dowd T, O’Kelly M, O’Kelly F. Structure of General Practice in Ireland 1982-2005. Dublin: Department of Public Health and Primary Care, Trinity College Dublin and Irish College of General Pr, 2006 http://www.medicine.tcd.ie/public_health_primary_care/research/rep orts/GP_Structure.pdf&#160;(Accessed 2 Aug 2007).
3. Allgar VL, Neal RD. Delays in the diagnosis of six cancers: analysis of data from the National Survey of NHS Patients: Cancer. Br J Cancer 2005; 92:1959-70.
4. Daly HM, Collins C. Early detection of cancer: A needs assessment of GPs.Dublin: Irish College of General Practitioners and Irish Cancer Society, 2006. http://www.irishcancer.ie/pdfs/Final_Cance_%20 Needs_Assessment_Report.pdf (Accessed 2 August 2007).
5. Gorman DR, Mackinnon H, Storrie M, Wilson GS, Parker S. The general practice perspective on cancer services in Lothian.Fam Pract 2002; 17: 323-328.
6. National Cancer Forum. A Strategy for Cancer Control in Ireland. Dublin: Department of Health and Children, The Stationery Office: Dublin, 2006. http://www.dohc.ie/publications/pdf/en_cancer_control_2006.pdf?di rect=1&#160; (Accessed 2 August 2007).
7. O’Sullivan P. Report on audit of GP Referral Guidelines for Suspected Cancer. Dublin: Health Services Executive East Coast Area, 2005.
8. National Institute for Health and Clinical Excellence (NICE) Referral guidelines for suspected cancer: clinical guidelines no. 27. London: National Institute for Health and Clinical Excellence, 2005 http://www.nice.org.uk/page.aspx?o=cg027 Accessed 2 August 2007).