GPs are dealing with more diagnostic tests than ever before. In the race to detect disease sooner and save more lives, access to those tests is being made easier.
Not only can GPs now order tests themselves, circumventing specialists and accessing cancer diagnostics on their own, but advancements in technology are creating greater volumes of testing.
Health secretary Matt Hancock’s long-term NHS plan pledges to ‘invest in new equipment, including CT and MRI scanners, to deliver faster and safer tests’. Mobile CT scanners in supermarket car parks and a new ‘digital diagnostic imaging service’ to boost image transfer to specialists are part of the vision.
Imaging and testing in general are snowballing
Dr Katherine Alsop
Improved diagnostic capability is a plus for patients wanting a quick and accurate diagnosis, but there is a problematic corollary for GPs: an increased appearance of incidentalomas.
These findings – where a scan shows something unrelated to the original enquiry – often mean GPs are left to pick up the pieces, as results are returned to them for investigation.
Bristol GP Dr Katherine Alsop, who chairs the RCGP’s optimal testing group, says: ‘Imaging and testing in general are snowballing.
‘There are lots of drivers for this, including defensive practice, patient pressure, and private screening, plus health special interest groups driving the agenda of their individual disease.’
This is creating a vicious circle in which patients expect GPs to carry out myriad lines of enquiry, which often lead to inconclusive results and much angst.
RCGP overdiagnosis group chair Dr Jane Wilcock says: ‘GPs are under pressure to improve early cancer diagnosis and this can only happen at the expense of many non-cancer patients being investigated.
We need to embed facts and figures around potential harms
Dr Jane Wilcock
‘The harm is that a number of people will end up in a medical system with non-serious illness and anxiety due to overdiagnosis of abnormality.’
She adds: ‘Increasing access and technology will only worsen the problem and we need to embed facts and figures around potential harms from inception rather than playing catch-up.’
No official data are collected on the prevalence of incidentalomas. But a systematic review of research published in the BMJ last year1 found some types of scan – including chest and colonoscopy CT – identified incidentalomas in more than a third of cases. The authors described incidentalomas as a ‘modern medical crisis’.
Pulse’s analysis of NHS England statistics shows an increase in the type of scans that reveal incidentalomas, as well as a sharp rise in the number of scans ordered directly by GPs themselves.
In 2018, there were around 40 million imaging tests reported in England, a 16% increase since 2013 – but the number of ultrasounds, CT scans and MRIs, which typically throw up incidental findings, has grown by 33%. ‘Direct access’ tests ordered by GPs have shot up by as much as 136% in some cases.
Full-body scans, also likely to pick up incidentalomas, are thought to have become more popular in recent years too as private companies push ‘full body MOTs’, although there are no official figures.
One in five GPs reports seeing incidentalomas on a weekly basis, according to a recent survey of around 950 GPs by Pulse. A third say it is a problem that has already been increasing in the past year, before implementation of the long-term plan, and is causing frustration – as GPs are often not given a steer as to how serious the unexpected findings are and whether further investigation will benefit the patient.
One GP, who asked to remain anonymous, sums up the frustration: ‘This is getting worse recently. We had a case of a missed diagnosis of a stroke that led to a lumbar spine MRI, which showed an incidental bone lesion. This required rescanning, was benign but showed up an ovarian lesion. Surgery followed, a GI bleed and ITU. Histology all benign.
Another GP adds: ‘I had a recent case in which a sonographer who identified a Bosniak [kidney] cyst made no comment on its significance. When I queried the grading and significance the reply I got was “That’s up to the GP to decide or refer to a renal consultant”.’
It is not always clear what a consultant has found in the first place. London GP Dr Anita Reddy says: ‘This week when I requested an abdominal X ray to try to locate a missing coil, the comment made was: “This patient may have a Riedel’s lobe”. I had no idea what this is or how significant it is.’
There are reasons behind the rise in imaging. It is most commonly used to detect signs of cancer and the NHS has made early diagnosis of cancer a key pledge.
Targets have been brought in to speed up diagnosis. The two-week wait from GP urgent referral to first consultant appointment for suspected cancer has been long established in England. And following a review by the independent cancer taskforce2, the NHS will now bring in a new target so patients receive a definitive diagnosis within 28 days.
Some patients cope fine with uncertainty and others literally become ill
Dr Katherine Alsop
Meanwhile, NICE guidelines from 20153– which said GPs should have direct and rapid access to tests for suspected cancer – have increased the pressure to order diagnostic imaging.
Taken together, these efforts will of course pick up more disease – but also more incidental findings.
On top of this are private healthcare groups offering more diagnostic tests and screening, and campaigning groups that are pushing more disease awareness. There is also pressure from patients and increasingly defensive practice by GPs concerned about complaints and investigations.
Glasgow GP Dr Margaret McCartney, a vocal champion of evidence-based medicine, insists the NHS should not be left to investigate incidental findings thrown up by private screening after the company has profited and walked away.
‘People should be given independent information about why these tests are not recommended by the NHS and my view is that the regulators should assess this and ensure that it is done,’ she says.
Both GPs and patients suffer when incidentalomas prompt further investigation. On the GP side, as RCGP optimal testing group chair Dr Alsop puts it: ‘GPs are sinking under increasing workload from many directions. This is one straw of many on the camel’s back.’
And there are concerns about the effects of incidental findings on patients. Dr Alsop says: ‘The trouble is some people cope fine with uncertainty and others worry themselves sick and literally become ill due to the abnormality they would never have known about. I try to do as few tests as is clinically safe because of this.’
We don’t really have enough quality evidence on rates and ideal follow-ups
Dr Margaret McCartney
While the profession agrees the way to deal with the situation is to advise patients of the possibility of incidentalomas from the outset, evidence is lacking, making it difficult to have meaningful conversations.
Dr McCartney says GPs ‘still don’t really know how to handle’ incidental findings – and how to raise the possibility when they first refer patients based on their original symptoms.
‘My view is that the original referring doctor should probably be telling patients about this possibility [of incidentalomas] but it’s hard to give good information as we don’t really have enough quality evidence on rates and ideal follow-ups,’ she says.
She advises that, particularly for head imaging – which often throws up incidental findings – patients should be routinely informed, especially due to the impact any findings might have on their life and work, and also on insurance.
When there are unexpected findings, it is not always clear who is responsible for following them up with patients.
Responsibility to act on findings rests with the clinician who ordered them
Dr Sam Finnikin
Sutton Coldfield GP Dr Sam Finnikin says the GP usually ends up deciding whether to follow up with a scan or refer on to a specialist.
‘Responsibility to act on findings rests with the clinician who ordered them. But in reality many will end up back with the GP, rightly or wrongly,’ he says.
Dr Finnikin says he prefers dealing with findings himself – rather than one consultant passing them to another – so he can take time to speak with the patient about their options, which he says consultants often don’t do.
‘Then I get the chance to talk to the patient about whether they want to have further investigations and risk overdiagnosis, or whether they’re happy to leave me carrying the small risk of missing something significant,’ he says.
However, some GPs think consultants should take a more active role.
Surrey GP Dr Martin Brunet says: ‘Where I think GPs should push back is that hospitals should be far better at communicating directly with their patients rather than sending information via a GP, or sending a technical letter to the GP and copying the patient in.
‘Hospitals need to know that patients want to know the result of scans and not be left in limbo for months waiting for a follow-up appointment.’
There is still a culture, perpetuated by policies and rhetoric from the top, that “more is better”
Dr Sam Finnikin
Consultants would also like to see more robust guidance on how to handle the problem. Dr Thomas Booth, a diagnostic and interventional neuroradiologist at King’s College Hospital NHS Foundation Trust in London, has long raised such concerns.
He says that, despite various attempts to produce guidelines on incidental findings, there is ‘no evidence to show we have got a handle on this’.
‘The concept of SPEW (scans propagating exponential workloads) is more real than ever,’ he says.
According to Mr Booth, every radiologist has a story of an initial finding turning out to be of ‘negligible clinical significance’ yet requiring ongoing management and follow-up.
But regardless of which part of the system deals with the findings, there is still a cost for the additional testing. Dr Brunet says: ‘There are the resource implications for both secondary and primary care.’ Direct GP access to tests will only make this worse, he believes.
RCGP overdiagnosis group chair Dr Jane Wilcock says evidence is urgently needed on the harm to patients put through the medical system with a non-serious illness.
‘All investigations should have clear side-effect and harm data for GPs and patients to discuss. As technology advances we need to embed this at inception.’
But for some, the problem lies with policymakers’ ‘more is better’ approach.
Dr Finnikin warns: ‘There’s a fallacy that more investigations will increase certainty, but in many cases they raise more questions than they answer.
‘There is still a culture, perpetuated by policies and rhetoric from the top, that “more is better”. This is harmful and causes pressure from patients for increasing investigations, and decreasing ability of patients and clinicians to deal with uncertainty.’
Common incidental findings
Around 90% of liver cysts are asymptomatic and only detected through ultrasound or CT scan for another reason. The vast majority will be of no consequence and decisions around whether follow-up is needed should be based on size or presence of any suspicious features on imaging.4
Between 30-50% of over-50s will have one or more renal cysts. These are largely benign, asymptomatic and only discovered by ultrasound scan for another issue. Size and appearance of cysts (in the absence of symptoms) should guide if more investigation is needed.5
The American College of Radiology recommends no further imaging for incidental thyroid nodules – because they are so common – unless they meet age-based size criteria, or there are other suspicious signs. Patients with no symptoms and no family history should only be referred for ultrasounds if nodules are bigger than 1cm for the under-35s or 1.5cm for the over-35s.6
Abnormalities on brain scans
A systematic review of patients with no neurological symptoms having brain MRI ‘health check-ups’ found 37 people needed to be scanned to detect one incidental finding. Researchers noted a lack of evidence about whether to follow up.7 A Dutch study of 2,000 middle-aged and older adults found incidental MRI findings in 10% – including meningiomas (2.5%), cerebral aneurysms (2.3%), arachnoid cysts (1.6%), and pituitary abnormalities (1.2%). Few required follow-up treatment.8
Incidental adrenal masses, found via ultrasound or CT scans, are benign in most cases. One study found they appear on 0.6-1.3% of abdominal CT scans. Even for large 6cm diameter masses, estimates show 60 surgeries are needed to identify one adrenal carcinoma. Symptoms – such as for Cushing’s syndrome – are important to determine if the mass is problematic.5
Identifiable in 1.5-4.5% of routine ultrasounds, gallbladder polyps are common.9,10 Gallstones are key factors in assessment. If over 2cm large the malignancy risk is high. But those under 1cm are 25 times less likely to be malignant than those greater than 1cm.11 Patients with gallstones or polyps over 2cm should be referred for surgery. Small polyps may need imaging (up to a year) to monitor changes.12
Lung nodules detected on CT are common but can be worrying for fear of lung cancer. One systematic review of CT screening lung cancer trials found a lung nodule was detected in up to 51% of participants. Yet clinical guidelines note more than 95% are benign. Age, smoking status and clinical characteristics of the nodule (size, location) should be considered when determining risk. 13
Most patients with ovarian cysts are asymptomatic and only discover them during ultrasound or pelvic exam. The Royal College of Obstetricians and Gynaecologists recommends postmenopausal women with asymptomatic, simple, unilateral, unilocular ovarian cysts less than 5cm in diameter have a low cancer risk. If serum CA125 levels are normal, cysts can be managed ‘conservatively with a repeat evaluation in 4–6 months’.14
1 O’Sullivan J et al. Prevalence and outcomes of incidental imaging findings. BMJ 2018;361:k2387
2 NHS England, 2015. Achieving World-Class Cancer Outcomes:A Strategy for England 2015-2020
3 NICE, 2015. Suspected cancer: recognition and referral. NG12
4 Gor R et al. Management of Incidental Liver Lesions on CT. J Am Coll Radiol 2017;14:1429-1437
5 Higgins J et al. Evaluation of Incidental Renal and Adrenal Masses. Am Fam Physician. 2001 Jan 15;63(2):288-295
6 American College of Radiology, 2017. Don’t recommend ultrasound for incidental thyroid nodules
7 Morris Z et al.Incidental findings on brain magnetic resonance imaging. BMJ 2009;339:b3016
8 Bos D et al. Prevalence, Clinical Management, and Natural Course of Incidental Findings on Brain MR Images. Online: Jun 23 2016
9 Heyder N, Gunter E, Giedl J, et al. Polypoid lesions of the gallbladder. Dtsch Med Wochenschr 1990; 115:243. 2
10 Jorgensen T, Jensen KH. Polyps in the gallbladder. A prevalence study. Scandinavian Journal of Gastroenterology 1990; 25:281
11 Youp Park et al. Long-term follow up of gallbladder polyps. Journal of Gastroenterol Hepatol. 2009; 24 (2):219-222
12 Homerton Hospital Radiology Guidelines for gallbladder polyp follow-up
13 Anderson I, Davis A. Incidental Pulmonary Nodules Detected on CT Images. JAMA. 2018;320(21):2260-2261
14 Royal College of Obstetricians and Gynaecologists. The Management of Ovarian Cysts in Postmenopausal Women. Green-top Guideline No. 34 July 2016