A few lessons in screening for Gordon Brown

Abstract

The prime minister seems just to want to offer more consumer choice for the worried well

Prime Minister Gordon Brown has just announced that he will allocate funding for a national screening programme concentrated on diabetes, cardiovascular and renal diseases, and stroke (BMJ 2008;336:62-3 doi:10.1136/bmj.39454.738912.4E). He said these were now seldom available except through private care but would, over the next three years, be made available to everyone.

The politicians, civil servants, and managers who devised this plan may believe it is evidence based. I’m not so sure. So far as I know, there has been only one controlled trial of such a policy. This compared age standardised mortality over five years associated with traditional demand-led care alone in Blaengwynfi, against similar care in Glyncorrwg, where this had been integrated with proactive screening and follow-up for chronic conditions (which included those listed by the prime minister) over the previous 20 years.¹ The two communities were socially almost identical. Both were in the bottom 5% of social ranking by Townsend index within the 55 local authorities then composing West Glamorgan. Apparent results were remarkable. Standardised mortality ratios for deaths at all ages were 36% above the UK mean in Blaengwynfi, 2% below it in Glyncorrwg. For deaths under age 65, the SMR was 60% above the UK mean in Blaengwynfi, 6% below it in Glyncorrwg.

Nearly an ideal experiment?

This was not an ideal experiment. The actual numbers of deaths under 65 were small—in Glyncorrwg, 21 deaths against 22 expected at UK rates; in Blaengwynfi, 48 deaths against 30 expected—so confidence intervals must be wide. The anticipatory care programme in Glyncorrwg evolved over 20 years, starting with blood pressure control and personal interventions on smoking, ending with a large bundle of interventions of many kinds, but all covering the entire population at risk. The population in Blaengwynfi was larger; medical staffing was smaller, with less continuity; and demand-led workload was higher (probably because anticipatory care eventually reduced contact rates).

However, it is hard to see how any such study could be more ideal than this. We were working in the real world, in an area of exceptional morbidity and mortality, with exceptional difficulties in recruiting and retaining medical staff, and accepting net incomes about half the UK average as a necessary price for good care. Looking at causes of death, the main differences lay in cardiovascular and respiratory disease, the pattern one would expect from effective early interventions.² So far as I know, nobody has tried to replicate it, but without such trials in the field, I think most clinicians in primary care remain unconvinced by evidence from much narrower, less sustained screening programmes, not integrated with ordinary demand-led patient care, mostly ignoring the logistic consequences of identifying massive additional workload, and almost all of them measuring surrogate outputs rather than deaths or health gain. So far, I believe our study is virtually all there is to justify any mass programme for anticipatory care for whole populations, which suggests that such programmes should pay at least some attention to what we did and how we did it.

An untrodden path

Our paper is rarely referred to, and I have never been consulted by anyone at the Department of Health since it was published. NHS Scotland has taken it seriously enough to invite me to Edinburgh to be lionised; the Glyncorrwg model of anticipatory care is cited in its 2005 plan for NHS Scotland,³ and on paper at least, is being pursued, but I have yet to see evidence that this has really modified clinical behaviour. It has probably had some influence in Wales, but it is hard to detect much of this on the ground, except perhaps in Rhondda-Cynon-Taff, the most energetic of our local health groups. For NHS England, where the grip of imposed consumerism on politicians, civil servants, and approved advisory bodies seems so tight that even David Cameron can find spaces through the left flank, I do not believe that anyone with power intends to follow the path we originated in Glyncorrwg.

What Gordon Brown probably has in mind, because it is all he knows, is screening as it has been developed as a commodity in the private sector: essentially a way for well people to buy reassurance, a great deal of it false or without evidence of benefit. For New Labour, the Big Idea is to make more accessible to poorer consumers whatever it is that richer consumers want, and thus broaden the market for its corporate friends—and shrink the NHS further toward its ultimately intended function as a brand name, and the undertaker of all things unprofitable yet unavoidably necessary.

Preconditions for success

Anticipatory care works. With all its limitations, our study showed that by tackling the hitherto undetected half of almost all important chronic disorders⁴ in an organised way at an early stage, integrated with demand-led care, there can be huge health dividends, and probably reduced loads on hospital care. If we ever get a UK government that listens to people experienced in the front line of health care, it will need to pay some attention to the following three preconditions for success.

Competition

Firstly, we have to remember that this is not new. Health promotion, screening, and dedicated clinics were placed on the GP menu by the Royal College of General Practitioners in 1981,⁵ and ever since July 1993 have been paid for, but despite a steep rise in recorded data there has been little overall effect on clinical decisions^{6 7}or morbidity,⁸ but a great deal of consequent cynicism both among doctors, and perhaps more still among the nurses and now health care assistants, to whom most of the work has been delegated.

Unless it is fully resourced (and this means more than simply giving money to GPs to spend as they wish), this work competes unsuccessfully with demand-led care, which is just as necessary and more obviously satisfying to both staff and patients.⁹ It always takes up staff time, but it seldom seems to engage much staff interest or intelligence, because the thinking appears already to have been done by the guideliners.

If the prime minister’s programme suggests anything new, it is that more of this work will be handed over to for-profit corporate providers, either competing with family doctors for this task or replacing them entirely for everything. As they are accustomed to running factories and supermarkets, they are expected to do this sort of work both profitably and economically.

Targetting

Secondly, to get a larger product from investment in anticipatory care we should target our sickest, most disadvantaged communities, in conditions of increasing polarisation of wealth that deregulated business can only increase: between 1930 and 1990 the gap in male mortality of social classes I and V increased almost two and a half times.¹⁰ Though this gross trend has levelled off since 1997 it has not reversed, and between 1995 and 1999, areas in the top 10% of Labour voting actually saw a rise in male mortality under 65.¹¹

Most health promotion activity is still concentrated in affluent areas with smaller, demand-led workload.¹² Almost certainly, the high yield in health gain in our study reflected the huge amount of unmet need in the South Wales valleys, typical of all areas where heavy industry and manufacture have been sacrificed to the new paper economy. Affluent areas have relatively little unmet need, and fewer difficulties in seeking it out or following it up; they will collect the data, but it may not make much difference to clinical workload, nor will it have much impact on either health or hospital referrals. But why would GPs in poor areas, already submerged under high, demand-led workload, welcome additional gratuitous burdens from well man and well woman clinics, other than to get the fees?⁷

The answer to this should be to face up to the larger staffing needs of poorer communities, for doctors as well as nurses and other support staff, and make NHS authorities, run by people experienced in health care and familiar with local staff and patients rather than business schools, give a clear lead. That would involve some painful changes in personnel which New Labour seems unlikely to contemplate; that’s not what they mean by painful choices. Instead, the prime minister will offer yet more consumer choice for the worried well to take themselves off to well person factories, coming off their assembly belts with recommendations for their GPs to implement, and congratulating themselves on their prudence compared with the feckless poor.

Integration

Thirdly, we have to integrate anticipatory care with demand-led care, so that screening procedures are applied imaginatively at times and in places where patients are already seeking help. We got exceptional compliance with treatment from our patients because they were given exceptional latitude to discuss whatever problems they wished. We knew them, and they knew us. We built everything on continuing friendly relationships with patients (including those whose problems made them unfriendly).

That is not what happens when GPs bolt health promotion and screening onto their demand-led care as an entirely separate operation, whose real value they frequently doubt, delegated to staff forbidden to make any clinical judgments but governed by guidelines generally untested in field conditions of primary care of whole populations, and derived from remote and often inappropriate sources.

Full population coverage

Admittedly, the very high contact rates (particularly with young men) typical of all heavy industrial and post-industrial areas gave us an advantage, even though it was associated with very high reactive workload. To get full population coverage we could rely on well over 60% of our population seeing a GP at least once a year, and over 85% of them within five years, giving us only 15% to contact proactively by phone, mail, or finally home visits if necessary. Our whole staff aimed at 100%, which is the ambition necessary for effective anticipatory care.

For the UK as a whole, contact rates are probably much lower, though the many breaches in the gatekeeper function of primary care deliberately created by government policy—and relentless and apparently unopposed growth self-referral to hospital accident and emergency departments—make this harder to measure. We also had a stable population with few in-migrants, in which it was relatively easy to accumulate rich profiles of personal data over many years and to form trusting relationships between staff and patients.

However, I know of practices in poor areas of London with annual turnovers of 30% or more which are successfully reaching all current targets for both data collection and subsequent treatment and follow-up; these have highly motivated staff. Interestingly, when these practices have recently had to compete for GP vacancies with large for-profit corporate providers, they have been unsuccessful. Disappointed applicants include a practice well known for its direct access to Brown’s predecessor, now presumably discarded.

Winning votes

One day somebody with power may understand that more votes might be won from continuity, real personal concern, and restored trust than from consumer choices where everything may be available except those three most precious and essential assets. This seems a long way off, but history is full of surprises.