A manager's lot would be greatly improved by a clear way of measuring outcome which shows how their actions impact on health, writes John Appleby

NHS managers have never had an easy time. Pressured from above and castigated from below, managers are simultaneously expected to implement government policy yet take the blame for policy failure.

If only managers could identify unambiguous measures of outcome for which they are directly responsible, and then show that their actions have improved such measures.What do managers actually do which could generally be recognised as any value to the NHS and its patients?

A recent study1 has asked just this question. Looking at management over the period 1991-92 to 1993-94, it concludes that 'overall, management input across and within hospitals does not appear to be associated with improved productivity as measured by average cost per adjusted inpatient episode'. The study found that increasing spending on top management was associated with lower levels of productivity, and decreased spending with higher levels. How can this be?

Before adding weight to the argument that the NHS is over-burdened with a grey-suited bureaucracy, it is worth noting some of the difficulties with this sort of study. The first is whether 'average cost per adjusted inpatient episode' properly captures the reason for managers' existence. If asked, hospital managers may well include some unit cost measure in their list of objectives, but what about waiting times, or quality of care, or implementing change with minimal disruption? The distance between managers (and their actions) and the ultimate healthcare outcome measure - health - together with the lack of knowledge of the connection (presuming one does exist) between the inputs (managers) and this outcome, poses a particular problem in terms of evaluating the contribution of management.

Managers may be very good at their (process) role of managing change and taking decisions, but they may make the wrong decisions or be stifled by factors outside their control. And effecting change can take longer than the three-year period of this study. Also, defining exactly who is a manager (the medical director? The medical records officer?) is problematic.

Despite these qualifications, few quantitative studies of this sort attempt to provide an evidence base which will help inform policy. Perhaps more than any group in the NHS, managers have suffered from policy apparently based on whim and political expediency. And although Soderlund's research may make disappointing reading for managers, much remains to be investigated about the connections between input, process and outcome in healthcare. A better understanding of these links, and the wider connections within the whole healthcare system, is in the long run the key to better policy- making - and a proper awareness of the value of management.