Intended for healthcare professionals

Rapid response to:


Measuring hospital clinical outcomes

BMJ 2013; 346 doi: (Published 30 January 2013) Cite this as: BMJ 2013;346:f620

Rapid Response:

Re: Measuring hospital clinical outcomes

I appreciate the rapid responses. I certainly agree with Dr. Greaves that we need to understand many dimensions of the patient experience. Events such as mortality or readmissions are just the beginning of what we need to understand about our performance and how patients experience our care is vitally important. There is much work progressing in patient-reported outcomes measures and we need to solve some of the challenges of implementing these measures in an efficient and meaningful way. When we speak about outcomes measures, we certainly include these outcomes within that scope.

As for language, please accept my apologies for not correcting our spelling – I really have no defence. I hope you understood the meaning and I do appreciate the opportunity to reach a British audience.

And as for disappointing the good surgeon Mahaffey, I did feel the need to balance a point of view with more technical information and I am sorry that you found our prose dry and our content lacking. I can assure you that merely counting events or describing experience will generally not suffice when differences have a modest magnitude as often occurs in the range of performance that we observe. To detect modest but meaningful differences – and to do so fairly – requires more sophisticated methods. Doing such measurement with the wrong tools would be akin to doing surgery with a dull knife.

And I agree that we need to draw attention to what is most important – but medicine and surgery are sufficiently complex that the array of measures will necessarily be large, though they exists in tiers of importance and should contribute to overall summary metrics. And I do believe that there is growing consensus about what measures should receive the most attention even as debate continues about data quality, statistical methods, and policy implications. I am optimistic because, despite our problems, medicine has never been more transparent about performance – and we are on a path toward even more transparency. And measurement has already sparked many tangible improvements in healthcare.

Competing interests: No competing interests

07 February 2013
Harlan Krumholz
Yale University
1 Church Street, New Haven, CT USA