Biases that make screening programs look good

Too often lung cancers are diagnosed at a late, untreatable stage, as was the cancer in the CT above. An effective lung cancer screening tool would be great!


Last week I described a recent trial that disappointed, failing to prove that lung cancer screening with low-dose CT (LDCT) saves lives. The designers of this trial are to be applauded for their trial design. They have understood, and tried to avoid, three forms of bias which can make a screening program looks like it is working - when it isn't. These biases have had an impact in the field of prostate and breast cancer screening also - where recent evidence suggests that benefits of screening in some populations (for example women aged 40-50) may have been overstated.

Imagine two people with very small, slow growing, early lung cancers attend different hospitals with indigestion, and are seen by different doctors who manage them independently. One doctor, on the barest provocation, arranges a CT chest for his patient and a small tumour is seen. The other patient does not undergo CT chest. After a run of investigations the first patient has his cancer removed and is followed up for 5 years before being declared free of disease. The second patient is discharged home, but 5 years later presents to his doctor coughing up blood - and a large lung cancer is found on chest xray. He too is treated surgically. Two years later, both patients present to their independent doctors with weight loss and abdominal pain. They are found to have distant spread of cancer and both die 6 months later. The first patient, it seems, survived 7.5 years from the time of diagnosis, the second only 2.5 years. Increasing survival from diagnosis for a lung cancer looks like a good thing, but in this instance survival may not really have been increased; both had cancer at time zero, but only one was diagnosed. This lead to an appearance of increased survival when the cancer was just diagnosed earlier. We call this lead time bias.

The second form of bias is like it. Suppose that a screening program only picks up slow-growing tumours. Suppose aggressive tumours generally make their presence felt clinically within about 6 months and are picked up clinically, and therefore not on screening CT. If a trial predominantly picks up slow-growing tumours then it is possible that a trial will seem to indicate increased survival time in patients undergoing screening only because they are the ones with all the slow-growing tumours. this is called 'length bias'.


Finally, lung cancer prevalence increases with age. It is likely that a number of older people die with lung cancer, rather than from it. This was certainly the finding in the recent Dante trial, the subject of the previous blog. A trial of LDCT screening may pick up, and 'cure', a number of indolent tumours that were never going to make their presence felt anyway. This is 'overdiagnosis bias'.

To overcome these biases it is very important that trials of screening programs are run in a randomised and controlled way. There must be a control group managed in 'usual' fashion. The results of a screening program cannot be compared to historical controls/cancer survival data. And they must be run over an adequate number of years to demonstrate if there really is an overall survival benefit. After all, a screening program that doesn't result in less people really dying from lung cancer, while it may serve to employ lots of people in medical jobs, is in essence a waste of time.

Andrew

Lung cancer screening

Why are we not screening for lung cancer?

So began an editorial in a recent edition of 'The Blue Journal'. In this particular edition, an Italian group was reporting on the DANTE trial, which investigated - in a randomised and controlled structure - the use of low dose CT scanning (LDCT) as a screening tool for lung cancer.

Now, although - as previously blogged - lung cancer is the leading cause of cancer death in our community, discussions about lung cancer screening don't rate a mention in the mainstream news. (Contrast this with prostate, breast, cervical and bowel cancer for example).

Efforts to determine whether our new, super-duper, multi-camera CT scanners can be utilised to screen for lung cancer have - to date - fallen over at the first hurdle. While they have demonstrated, repeatedly, that if you perform CT scans on lots of middle-aged smokers you identify lots of asymptomatic, early, curable tumours they have not yet proven that taking out these tumours saves lives.

The recent study has struggled at the same point.

What did they do? A total of 2,472 men aged 60 to 75 years, with at least a 20 pkt year smoking history (a packet a day for 20 years equivalent) were randomised to either a baseline LDCT scan and annual LDCT scans over 5 years, or else a baseline chest xray with annual clinical review. The current data cover follow up over a mean of 33 months (ie not a very long time yet).

What did they find? They found 60 patients (4.70%) with lung cancer in the LDCT group (63 cancers in total - some patients had more than one). Twenty-eight cancers were found in the initial scan ( ie over 2% of these people had cancers they didn't know about to begin with ), 25 were found on subsequent scans and ten were found because people developed symptoms between their annual scans. In the control group, they found 36 cancers in 34 patients; 8 at basline, 3 at routine review and the others because patients developed symptoms.

What did they do next? Lots of intervention! Thirty-nine of the 60 patients in the LDCT arm with cancer had resections, and of these 36 were complete resection. In the control group it was 18 of 34 resected, 17 completely. In the LDCT group there were 96 invasive procedures, vs 36 in the control group. As the investigators put it ''we performed three times as many invasive procedures and found twice as many lung cancer cases with LDCT than without it, ...(but) .. the absolute numbers of advanced and lethal lung cancer cases were unfortunately identical in the two arms.....moreover a significant proportion of major surgical procedures (13%) were performed for pulmonary lesions that ultimately turned out to be benign..."

What happened next after all that intervention? Mortality rates in the groups are identical to date. Put in plain English, the same number of people died in each arm of the study.

So, LDCT screening for lung cancer in asymptomatic patients with no history of cancer detects nearly twice as many cancers as a 'chest xray and annual specialist review' approach, leads to a raft of invasive and expensive interventions, and has not been proven to save any lives. Yet.

Andrew

Screening for obstructive sleep apnoea

It took me a little while to get around to opening Sleep Journal from the start of last month, but I have enjoyed reading a study from Japan which validated a simple questionnaire intended to screen a community-based population for obstructive sleep apnoea.

This study group (Sleep Vol 32 No 7 p939-948) created a questionnaire that required only information about a person's gender, body-mass index, snoring history (yes or no) and blood pressure. From this information a score out of 18 was obtained. Scores above 11 were increasingly predictive of the presence of occult obstructive sleep apnoea.

Why did I like this paper?

Firstly, there was no question about sleepiness. Realising that people in the community who have obstructive sleep apnoea may be differentiated from those who go to the doctor with the problem by the absence of sleepiness, this somewhat-nebulous symptom was left out of the mix.

Secondly, a neat combination of two investigations that fall short of a full sleep study was used to evaluate for the presence of obstructive sleep apnoea. A home monitoring device was used which measured only the usual respiratory channels that we measure in the sleep lab - air pressure changes at the nose, thoracic and abdominal elasticated bands as well as pulse oximetery But how did they know when the person was asleep? A wrist actigraph was worn. This is essentially a motion sensor, the size of a wrist watch. When motion stops, sleep is assumed. This sort of device is used in evaluation of people with problems such as shift work sleep disorder, as it is more objective then a sleep diary. The combination of actigraphy and respiratory monitoring is much more simple then a full sleep study. Together they constitute, I think, reasonable way to follow up a community screening questionnaire. There will probably be more of this sort of diagnostic evaluation performed here once medicare agrees to pay for such a test.

Finally, I think we could use something like this here. It should not be difficult to validate for local conditions (our at-risk BMI is likely to be higher than the Japanese). And that has got me thinking ....