A reply

We agree with Siegmueller and Herden-Kirchoff’s cardinal suggestion, namely that it is necessary to define efficiency/productivity in financial terms. Our papers did not do this. Early attempts have been thwarted because, although tariffs (the fees paid to the hospital for a procedure) are known, good data on costs (which must include staff costs, consumables, drugs, equipment, theatre running costs, etc.) for all operations are very difficult to ascertain [1, 2]. Service-line reporting is still in its infancy and it remains possible that high operating theatre costs with certain procedures may be counterbalanced by, for example, shorter inpatient stays. We sincerely hope that Siegmueller and Herden-Kirchoff will develop their interest in our work by testing our predictions of efficiency/productivity against robust costing data. Siegmueller and Herden-Kirchoff allude to some economic definitions of efficiency/productivity but in fact there is no single definition. Economists subdivide efficiency into context-specific terms such as ‘X-efficiency’ [3], ‘allocative efficiency’ [3], Pareto efficiency [4], Kaldor-Hicks efficiency [5], and so on. The Organisation for Economic Co-operation and Development therefore stresses that each discipline needs to develop its own measures [6]. Since operating theatres ‘produce’ nothing it follows that productivity and efficiency can only be assessed by other means such as use of time and ability to complete the work scheduled (akin to antique clock restorers or car repair shops [1]). We are pleased therefore that Siegmueller and Herden-Kirchoff ‘agree with our conclusions and state that our analysis is ‘a useful starting point’ in these important questions. Yet they may have failed to appreciate a subtle yet powerful application of our analyses. Our formulae really describe efficiency and productivity in measures of activity or work input; i.e. the factors under direct control of those who work in operating theatres. We did not measure the value assigned to this work by those outside theatres. By analogy a coal miner and stockbroker might work equally effectively in their working day, but society assigns higher fiscal value to the efforts of the latter than it does to the former. Thus in purely monetary terms, even an unproductive stockbroker performs better than the most productive miner. The best that individuals in theatre can hope to do is utilise all their allocated time, work as fast as reasonable, not over-run, and not cancel patients. They can do no more. Therefore a rational and ideal funding system should reward these teams. If instead, subsequent costing analysis shows such teams to be less profitable than teams that waste time, over-run, or cancel patients, then this suggests some perverse incentives in the system of remuneration (i.e. the fault is likely to lie with the funding method, not with our formulae). Furthermore, if two teams are judged equally efficient or productive by our formulae, yet one is more profitable than the other, then this suggests that some things other than utilisation/cancellation/team speed are being fiscally valued, and it would become important to establish what these things might be. The relevant questions for society or the health service might then turn out to be those such as: should neurosurgery (like stockbroking) be inherently more profitable than gynaecology (like coal mining) or vice versa, regardless of their respective cancellation rates, utilisation, etc.? In other words: what should we value? Therefore focussing on costs alone without applying our analyses, as Siegmueller and Herden-Kirchoff imply is desirable, would simply identify the profitable teams. These teams should ideally be the same as those identified as efficient/productive by our formulae. Any inconsistency between the two analyses will help identify teams that are profitable purely through perverse incentives. It is the existence of perverse incentives, rather than inefficiency, that might be the main problem facing the National Health Service [7].