The Maynard Doctrine – IT’s an evidence thing

Professor Alan Maynard OBE appraises the cost-benefit possibilities – expected and otherwise – of computerisation in healthcare

We book our travel, hotels and holidays electronically. We access our bank accounts and shop electronically. And we are continually told that information technology is about to revolutionise the NHS.

So why is the revolution such a long time coming?

The first problem is the chaps and chapesses who get paid over £100,000 per year to work in the NHS. There are two overlapping types of such persons.

”Non-clinical managers who generally have all too little training in analytical methods, let alone IT … increasingly live in a world flooded with data, yet one in which there is little information suitable for enhancing the quality of their management”

Firstly there are the non-clinical managers who generally have all too little training in analytical methods, let alone IT. Increasingly, they live in a world flooded with data, yet one in which there is little information suitable for enhancing the quality of their management.

`Fear of IT
They are intimidated by IT. They know they need it to ride out the depression and they suspect that IT, substituting capital for people, may be their salvation. But they use IT very cautiously as it may fail and thus will make their life even more complicated, particularly as it will require then to be more strategic and less tactical.

The second group of chaps and chapesses who influence the use of IT go around “bare below the elbows” and stripped of their watches, jewellery and ties. These folk have been trained to act with clinical autonomy, but need nowadays to work in teams.

Co-ordinating and performance managing such teams requires integrated real-time information about patient flows and adherence to practice standards. The scope for IT seems enormous for these folk; but practical delivery is so slow coming. In part this is a product of their reluctance to let anyone but themselves design a support system, because their needs are “unique” and systems may erode their autonomy - at long last!

Evidence needed of IT improvement
Enthusiasts for IT in the NHS continually press for more investment, but continuing to simply throw money at IT in the belief that failure to spend is the problem, would be unethical.

"IT investment should be conditional on evidence of cost-effectiveness. In the absence of such evidence, IT should be treated like new treatments: as experimental"

IT investment should be conditional on evidence of cost-effectiveness. In the absence of such evidence, IT should be treated like new treatments: as experimental. As such, IT should be piloted using rigorous methods of evaluation, either randomised controlled trails or evaluations of costs and effects that use quasi-experimental methods of appraisal.

Such practices are of course alien in healthcare systems where managers and clinicians know things work in the absence of evidence and throw resources around in ways which they would never use if they were spending their own budgets.

Evaluation begins (overseas)
But all is not lost! To the embarrassment of these cavaliers and wastrels, more careful folk are now beginning to evaluate investments in IT with more care.

One recent study analysed the use of IT by doctors and its apparent effects on medical errors (Amarasingham et al 2006, Archives in Internal Medicine January 26th, 2009. They devised a system to measure the use of IT by clinicians in four areas: notes and records; test results; order entry; and clinical decision support. They conclude that those practitioners and institutions which used such IT systems more had lower levels of medical errors.

“If you can get clinicians to use IT support systems, you are less likely to damage your patients”

This success in reducing damage to patients could of course be a product of other factors - correlation does not prove causation. The authors of this Texan study sought to adjust their data to take account of such compounding factors as the affluence of individual institutions. To the extent that they were able to adjust for such biases, the results appeared to be robust. Thus if you can get clinicians to use IT support systems, you are less likely to damage your patients

Informing consent
Another use of IT that appears to be promising is patient information systems. Clinicians use decision-making criteria to determine whether a patient would be eligible for a hip or knee replacement, or for a prostate operation. Take for instance a man who gets up many times in the night to urinate, with significant disruption of his sleep pattern: should he be operated on by your friendly neighbourhood plumber-urological surgeon?

It is clear that if the patient follows the clinician’s advice, surgery would follow. However if you explain, using a nice PC and simple programming, that surgery is risky, the patient may decline surgical intervention.

Surgery may not mitigate incontinence. Indeed it may make it worse and leave you using a nappy for life. Furthermore, the surgery may destroy your sexual functioning. Faced with such risks, many a man prefers, to continue to get up in the night for a pee!

Similar conservatism is exhibited by patients offered hip and knee replacements. A Canadian study analysed the identification of needy patients using clinical criteria and then asked them did they want it. They found that many respondents, when made aware of the problems of surgery, decided to wait. (Hawker et al, Medical Care, 2001)

This nice disjunction between clinical need and patient demand is of interest to all those managing waiting lists. Perhaps there is unnecessary surgery being undertaken. If patient preferences were better informed by simple IT, might the demand for some elective procedures be reduced?

IT has considerable promise to improve processes and outcomes of care. The tradition of investing without evaluation is common to medicine and has wasted considerable resource. As studies begin to emerge about the effectiveness of IT, more efficient investments may result.

The potential of IT to improve the quality of patient care and improve efficiency is considerable. But caution is the message; and IT enthusiasts should always be challenged to produce evidence rather than noble intentions and vacuous assertions, which all too often waste resources.