Cancer referring in general practice

Background

Over 250 000 people in England are diagnosed with cancer each year, and 130 000 suffer cancer-related deaths annually (Department of Health, 2012). With ever-increasing financial pressure on the NHS, correctly diagnosing, referring and allocating resources is essential. GPs see, on average, between eight and nine new cancer patients per year. However, many more patients have a clinical picture indicative of cancer, thus complicating the decision to refer to secondary services (National Cancer Intelligence Network, 2010). The NHS Cancer Plan in 2000 demonstrated dedication to improving cancer outcomes by implementing a maximum 2-week wait (2ww) for all cancer referrals (NHS, 2000). On average a GP will make around 25 2ww referrals annually (Department of Health, 2011).

Firdale Medical has a practice population of 7868 patients and a 2ww cancer detection rate of 7.03%, which is the lowest in the Vale Royal Clinical Commissioning Group (CCG) (NHS Vale Royal CCG, 2011). A four-fold variation in 2ww referral numbers across Primary Care Trusts was recently reported, highlighting the need for a closer look at the nature of referring practice and what factors affect the decision to refer (National Audit Office, 2010).

The primary rationale behind this audit was to improve the standard of 2ww referrals to provide earlier diagnosis and access to treatment. The audit sought to focus on existing National Institute for Health and Care Excellence (NICE) referral guidelines and adherence to them within the practice, accepting that the decision to refer is multifaceted and a minority of referrals may not fit this model. Additionally, the audit sought to reduce expenditure by reducing the number of referrals to secondary care deemed inappropriate.

Standard

The audit sought to meet the following standards:

100% of 2ww referral clinical reasoning deemed appropriate as per NICE guidelines and case-by-case judgement where appropriate (NICE, 2005)

Design

A prospective audit of 2ww referrals to secondary care by eight GPs from Firdale Medical was performed. The first cycle between 1 January and 31 May 2013 was represented by 68 patients (30 male, 38 female), with ages ranging from 21 to 88 years (median age: 63.7 years). The second cycle was from 1 June to 31 October 2013 and was represented by 64 patients (23 male, 41 female), with ages ranging from 19 to 90 years (median age: 62.4 years).

Data was collected from the 2ww referral paperwork, medical records and hospital correspondence. Data for standards 2 and 3 were quantifiable and obtained directly from the above sources. Standard 1 was subjective, so a partner GP at the practice was consulted to advise if referral was appropriate. All identifying information was removed and any documented clinical reasoning in the notes was used to support referrals when criteria had not been met. Consequently, inappropriate referrals refer to a failure to adhere to the 2ww criteria for each individual referral speciality, accompanied by a lack of clinical reasoning in the notes to justify urgent 2ww referral.

Audit results

The results for cycle1 can be found in Table 1. The three exclusions were due to a patient’s outcome not returning within the time frame and two cancelled appointments. The cancer detection rate was 21.5%. This is significantly higher than the 7.03% standard previously achieved by the practice and only 2.31% lower than the practice with the best yield within the CCG. Despite progress with detection rates, some speciality referrals were consistently deemed to be more inappropriate than others. These included dermatology (29%) and breast (35%), highlighting the difficulty in referral decisions in these areas. However, the most commonly detected cancers were squamous cell carcinomas (SCCs) and breast cancers. A breakdown of inappropriate referrals by referring route can be seen in Fig. 1.

OPEN IN VIEWER

Table 1.

Audit results.

Standards		Cycle 1
		Yes	Percentage (%)	No	Percentage (%)	Achieved?
1	100%	48	73.8	17	26.2	No
2	100%	48	73.8	17	26.2	No
3	17%	14	21.5	51	78.5	Yes

Standards		Cycle 2
		Yes	Percentage (%)	No	Percentage (%)	Achieved?
1	100%	48	75	16	25	Improved
2	100%	60	93.8	4	6.2	Improved
3	17%	19	29.6	45	70.3	Yes

OPEN IN VIEWER

Figure 1.

Cycle 1 percentage of inappropriate referrals by 2ww referral route.

There was diversity in the number of referrals made by each GP (range 0–18) and the most referrals were made by the GP registrar; 26.2% of referrals made should have been sent down alternative referral lines and only 73.8% of referrals were made within one working day (range 1–10 days).

Intervention

The following actions were communicated to the GPs and implemented immediately:

Referral uncertainty is to be discussed at the daily referrals meeting to allow a decision to be made and the referral sent within one working day

Re-audit results

Using the cycle 1 audit template, the results for the second cycle include 64 referrals with no exclusions, demonstrated in Table 1. In the re-audit, there remains diversity in the number of referrals made by the GPs (range 5–17) but it was found that 93.8% of 2ww referrals were now sent within one working day: a 20% improvement on cycle 1.

An improved cancer detection rate of 29.6% is reported; an increase of 8.1% from the first cycle shown in Fig. 2. The total number of inappropriate referrals has reduced by 1.2% from the first cycle. Notably, the breast cancer detection rate remained consistent but inappropriate breast referrals decreased by 28.9% following the interventions. Inappropriate lymphadenopathy referrals also decreased by 11.8% in cycle 2. Conversely, Fig. 3 shows that inappropriate dermatology referrals remain a problem as they increased by 8.1% despite the interventions as did urology and ENT referrals, by 12.7 and 12.5%, respectively. Changes among the other specialties were less significant due to low referral numbers. OPEN IN VIEWER

OPEN IN VIEWER

Figure 3.

Comparing the percentage of inappropriate referrals by 2ww referral route between audit cycles.

Discussion

The cancer detection rates reported within the audit are significantly higher than the practice had previously achieved; they are above the CCG 17% average and the national average of 11%. Breast cancers and SCCs were the most commonly found cancers and yet the most inappropriately referred for routes further complicating the decision to refer. The number of inappropriate referrals most notably reduced for suspected breast cancer in cycle 2, highlighting the impact that frequent revision of day-to-day practice can have on this problem. However, urology and ENT inappropriate referrals increased in cycle 2. Proposed explanations include: non-specific presentations, clinical inexperience in these specialities and a lack of diagnostics for suspected ENT malignancies outside of the 2ww system.

Overall, inappropriate referral could have been avoided in some cases by paying closer attention to the referring criteria, as three suspected basal cell carcinomas (BCCs) were referred via the 2ww dermatology pathway, despite guidance not to refer urgently for BCCs. For 2ww breast referrals, many were referred despite no lump being felt in cycle 1, or with non-specific symptoms that were deemed inappropriate for urgent investigation. Undeniably, a focus for future practice will be improvements in the knowledge base and confidence among the GPs with dermatology referring given the consistent number of inappropriate referrals during the audit. Engagement in dermatological education could help to eradicate inconsistencies with choice of referral route and help to standardise practice among the GPs.

In cycle 1 the most referrals were made by the GP registrar. Inexperience could account for higher numbers of referrals; however, it is equally viable that inexperience may lead to under-referral, supporting trainees to make the difficult referral decisions is therefore of paramount importance. Only 84.4% of all referrals were seen in a specialist clinic within the advised 14 days. Although not within the direct control of the practice, this highlights the burden on secondary care due to an inundated 2ww pathway and consequently the benefits of appropriate referral and consistent one working day referral.

Cycle 2 results demonstrate that almost all referrals were made within one working day of consultation and the overall number of inappropriate referrals had reduced. That said, the successes of achieving higher detection rates at the practice must be considered in the context of what constitutes an appropriate referral. High detection rates could be attributed to patients presenting with more advanced symptoms and consequently result in poorer prognoses. The currently discouraged act of referring for non-specific symptoms may not adhere to guidelines, may cause over-referral, lower detection rates and be deemed inappropriate as per this model; however, early referral before obvious signs are apparent may provide early diagnosis and access to treatment for cancers with historically late presentation and poor prognoses. Consequently, it is important to note that the decision as to the appropriateness of a referral is a contentious and highly topical one within primary care and cannot solely be accounted for by adherence to current guidelines alone.