‘Blood in pee’ campaign: Increased demand on secondary care with no change in cancers diagnosed

Objective: As part of the national Be Clear on Cancer campaign, the ‘blood in pee’ campaign was launched in 2013. We aimed to evaluate the impact of the campaign on 2-week wait (2WW) referrals and the resulting diagnoses of malignancy at a single trust, and secondly, to evaluate the socio-economic background of patients referred. Patients and methods: Suspected cancer 2WW patients in the 3 months pre- and post-campaign were included. Demographics, investigations and diagnoses were recorded. A Kolmogorov–Smirnov test demonstrated a normal distribution. The data were treated as parametric and analysed with the unpaired Student’s t-test. Results: Referrals for visible haematuria significantly increased by 52% from 135 pre-campaign to 205 post-campaign (p = 0.03). There was a fall in the proportion of patients diagnosed with malignancy from 20.27% pre-campaign to 15.36% post-campaign. The mean index of multiple deprivation score of referrals did not change: p = 0.43. Conclusion: This campaign has increased referrals without increasing the proportion of malignancies diagnosed, placing large demand on services without benefit or extra funding. Nor has the campaign effectively reached deprived socio-economic groups. There is little evidence as to the efficacy of untargeted cancer awareness campaigns and further work is needed to improve their pick-up of malignancies. Level of evidence: 2C


Introduction
About a quarter of new cases of bladder cancer diagnosed are at a late stage, a concern addressed by NICE in their guidelines highlighting symptoms associated with bladder cancer and the appropriate referral pathway for patients with suspected bladder cancer. (1) Several public awareness studies have shown that knowledge of symptoms and risk factors for cancer among the general population is poor (2). It is thought that increased awareness of key symptoms of bladder cancer would help earlier detection and improve survival. It is estimated that between 5,000 to 10,000 cancer deaths within 5 years of diagnosis could be avoided each year in England if efforts to promote earlier diagnosis were to be successful (3). In response, the UK Department of Health published "Improving outcomes: A Strategy for Cancer" in 2011. This strategy included public awareness campaigns designed to help people recognise symptoms of concern.
The national "blood in pee" campaign ran from October to November 2013 and was repeated in 2014 and 2016. The key message was: "If you notice blood in your pee, tell your doctor." The advertising highlights that finding cancer earlier makes cure more likely. The campaign was aimed at men and women above the age of 50 from lower socio-economic groups. This included advertising on television and radio, in newspapers, internet and branded pharmacy bags.
Whilst the 'Be clear on cancer' campaign was rolled out across several cancer sites, there have been reports that the campaign did not increase referrals from socio-economic backgrounds where poorer outcomes are more prevalent, and increased referral of the 'worried well', having no effect on detection rates (4)(5).
The aim of the present study was to evaluate impact of the campaign on service and diagnosis in a UK centre, and to see whether the campaign reached the targeted the socio-economic background.

Method
All patients referred to the Royal Derby Hospital using the suspected genitourinary tract cancer 2WW pathway 3 months before (June 2013 to August 2013) and 3 months after (November 2013 to January 2014) the initial campaign were identified using the hospital's cancer audit prospectively maintained database. Basic patient demographic data, clinical reason for referral, subsequent investigations, cost analysis and eventual diagnosis were all collected. Patients' postcodes were cross referenced with Indices of Multiple Deprivation (IMD) 2010 scores to identify socio-economic group of the patient (6). The stage of the tumour was also recorded if a diagnosis of bladder cancer was made. Costs were obtained from the hospital's patient level information and costing systems (PLICS). The appropriateness of referrals according to NICE criteria was not assessed.

Data Analysis
A Kolmogorov-Smirnov test was performed on the data, demonstrating a normal distribution. The data were therefore treated as parametric. Data was analysed with the unpaired t-test using Microsoft Excel 2013 (Microsoft, Seattle, USA).

Pre-campaign Demographics
Three hundred and sixty five patients were referred via the 2WW pathway in the 3 months prior to the launch of the national 'blood in pee' campaign. The median age was 66 years (range 19-98) and 73% were men (n=265). The number of people above the age of 50 was 326 (89%). A total of 239 (65%) referrals were for haematuria; both visible and non-visible (135 and 104 respectively). Ninetythree (25.4%) were referred because of elevated age-specific PSA and the rest were due to testicular lump(n=4), haematospermia (n=1), abnormal penile examination (n=11) or urinary tract infection with haematuria (n=7)( Table 1).

Post-campaign Demographics
Four hundred and sixty-one patients were referred via the 2WW pathway in the 3 months following the campaign. The median age was unchanged from pre-campaign at 66 years (range 18-92) and 68% were men (n=313). The number of people above the age of 50 was 414 (90%). A total of 327 (70.8%) referrals were for haematuria; both visible and nonvisible (205 and 122 respectively). One hundred and two patients (22.1%) were referred because of elevated age-specific PSA and 32 patients were referred for other reasons. (Table 1).
The total number of referrals for all reasons increased by 26% and referrals for visible-haematuria increased significantly by 52% (p=0.03). NHS England data on GP referrals for suspected urinary tract cancer were analysed for both periods to exclude an effect of seasonal variation on referrals. No effect was found (p=0.12)(7).
Table1. Symptoms triggering a 2WW referral for suspected urothelial cancer before and after the first national blood in pee campaign.

Reason for referral
Pre-campaign (n) Post-campaign (n)

Socio-economic factors
The average Index of Measures of Deprivation (IMD) 2010 score before the campaign was 17.09 (range 3.40-65.14) while the average IMD 2010 score for patients after the campaign was 17.9 (range 2.71-65.14). (p=0.43). The difference between socio-economic groups was not statistically significant despite the specific stated intention of campaign to target people from lower socioeconomic class.  Figure 1. In addition, 7 (5.2%) patients had a diagnosis of renal/ureteric cancer before campaign while 10 (4.6%) patients were diagnosed with same after the campaign. Two hundred and fourteen patients before and two hundred and eighty-nine patients after the campaign had no cancer diagnosed.  Figure 2 and Table 3 show the numbers of investigations performed and associated costs before and after the campaign for all haematuria referrals. Both numbers of investigations and costs increased in the period after the campaign, corresponding to the increase in referrals. £36,212 extra was spent on investigations after the campaign compared to before. Although the total cost increased postcampaign the cost per cancer diagnosed dropped from £3,528 pre-campaign to £3,198 postcampaign, however this is based on a non-significant change in number of cancers diagnosed.

Discussion
This is the first study to report the impact of the 'Blood in Pee' campaign on costs to secondary care. It is also the first to assess the socio-economic group of those patients referred.
We show that the Be clear on cancer "Blood in Pee" campaign significantly increased the number of suspected new cancer referrals with haematuria, but there was no significant change in the diagnosis of bladder or renal tract cancer leading to a drop in proportion of cancers diagnosed. Our In addition, the average Index Measure of Deprivation before and after campaign for both visible and nonvisible haematuria remained unchanged, indicating that the campaign had no effect on its demographic target. Archie et al reported a 92% increase in haematuria referrals with no corresponding significant change in cancer diagnosis in their analysis of the impact of the campaign on a single trust, agreeing with our findings (8).
These findings are in line with examination of the impact of the 'Blood in Poo' campaign for colorectal cancer, again demonstrating increased referral numbers with no change in cancer diagnoses and no change in the socio-economic background of those referred (4). However, Public Health England reported an overall increase in cancer detection rate from pilot studies (9), something not found in our data.
The campaign met its stated objective of increasing the number of people being referred for visible haematuria which increased significantly by over a half. Referral for nonvisible haematuria also increased by a more modest 17.3%. This increase in visible haematuria referrals would imply an increased awareness of the signs of urological cancer, and thus success of the mass educational aspect of the campaign.
While the criteria for referral for visible haematuria are straight forward, the criteria for nonvisible haematuria are slightly more complex and the increase would imply an increase in primary physician awareness about the haematuria referral guidelines, possibly increased furthermore by exposure to the campaign (leaflets in the workplace, screensavers on work computers etc) rather than patients themselves seeking attention, non-visible haematuria not alerting patients to a problem by definition. We did not however formally audit weather the patient and/or primary physician was aware of the campaign at time of diagnosis, which if collected may have allowed us to draw a more 'cause and effect' conclusion.
The main limitation to this study is that it represents a single centre experience, albeit with a large catchment population of 600,000. The Bladder cancer arm of this campaign is no longer live however other cancers are continuing to be targeted underneath the same campaign title of 'Be Clear on Cancer' and thus ongoing analysis of these campaigns is worthwhile.

Conclusions
In conclusion, while visible haematuria is considered a reliable predictor of underlying urological malignancy and while this mass media campaign did increase the number of referrals for visible haematuria, it failed to achieve its objective of increasing referrals from lower socio-economic groups and has also failed to significantly increase in the number of urothelial and renal cancers diagnosed. There is little evidence as to the efficacy of current untargeted cancer awareness campaigns and further work is needed in order to improve their pick up of undiagnosed malignancies.