Problems of measuring hospital output by case mix

To pay a hospital fairly, or to compare one with another, you need to know what it produces — and a hospital’s output is not beds or admissions but patients treated, each different. In 1986 Gwyn and colleagues were among the first to try the American answer, diagnosis-related groups, on NHS discharge data, and found the statistics and costings wanting. Hip fractures gave a test case for what medical audit and resource management could learn from each other. “Ways of seeing” steps back to ask why the same variations in hospital use are read in America as a measure of efficiency and in England as a measure of need.